VPLEX Administration Student Guide

VPLEX ADMINISTRATION
DOWNLOADABLE CONTENT
Internal Use - Confidential
Table of Contents
VPLEX Ecosystem ............................................................................................................... 8

VPLEX at a Glance .............................................................................................................. 9
Data Mobility Challenges ................................................................................................... 11
Active-Active Data Center with VPLEX .............................................................................. 12
Availability with VPLEX ...................................................................................................... 13
VPLEX Product Family....................................................................................................... 14
VPLEX for ALL Flash ......................................................................................................... 15
Where and When to Use VPLEX ....................................................................................... 16
VPLEX Use Case - Online Tech Refresh ........................................................................... 17
VPLEX Use Case - Storage Pooling .................................................................................. 18
VPLEX Continuous Availability........................................................................................... 19
Stretched VMware Clusters ............................................................................................... 20
VPLEX with RecoverPoint .................................................................................................. 21
MetroPoint Solution............................................................................................................ 22
VPLEX Architecture ................................................................................................. 23

Dell EMC VPLEX Logical Architecture - Director I/O Stack ................................................ 23
VPLEX Constructs ............................................................................................................. 25
VPLEX Local Architecture .................................................................................................. 26
VPLEX Metro Architecture Builds on VPLEX Local ............................................................ 27
VPLEX Physical Architecture ............................................................................................. 28
External Comparison of VS2 and VS6 Engines .................................................................. 30
Intra-Cluster Communications (Local COM) ....................................................................... 33
VPLEX Local Rack Configuration....................................................................................... 35
VPLEX Engine I/O Modules .............................................................................................. 36
WAN COM Performance Considerations ........................................................................... 38
VPLEX Power Subsystem .................................................................................................. 39
VPLEX Engine Battery Backup .......................................................................................... 40
VPLEX Management Server .............................................................................................. 41
VPLEX VS6 MMCS IP Connections ................................................................................... 43
VPLEX VS6 Management – MMCS ................................................................................... 44
VPLEX Administration-SSP

Page ii © Copyright 2020 Dell Inc.
VPLEX VS6 Management - MM ......................................................................................... 45
VPLEX Management IP Infrastructure ............................................................................... 46
VPLEX IO Operations............................................................................................... 48
Distributed Cache Coherency ............................................................................................ 48
VPLEX I/O Operations ....................................................................................................... 49
Path Redundancy .............................................................................................................. 53
VPLEX Cluster Witness ..................................................................................................... 55
VPLEX Management Options .................................................................................. 56

VPLEX Management Tools ................................................................................................ 56
VPLEX Management Security ............................................................................................ 57
VPLEX CLI Context Tree (partial listing) ............................................................................ 58
Key Management CLI Root (/) Sub-contexts ...................................................................... 59
Navigating the CLI Context Tree ........................................................................................ 60
Locating in the Context Tree .............................................................................................. 61
Dell EMC Unisphere for VPLEX ......................................................................................... 62
VPLEX Online Support ...................................................................................................... 63
VPLEX Connectivity ........................................................................................................... 64
Back-End Array Zoning Best Practices - Unity ................................................................... 64
Back-End Zoning - Quad VPLEX with VMAX ..................................................................... 65
Dell EMC XtremIO with VPLEX Configuration .................................................................... 66
Determine WWNs of VPLEX FE Ports ............................................................................... 67
Host Connectivity to VPLEX Front-End Ports ..................................................................... 68
FE Connectivity Best Practices .......................................................................................... 69
VPLEX Front End Connectivity .......................................................................................... 70
VPLEX Storage Provisioning Concepts ................................................................. 71

Storage View: Establish Host-to-Virtual-Storage Connectivity ............................................ 71
VPLEX Virtual Storage Constructs – Complex Device Example......................................... 72
View Virtual Volume ID Presented to Host ......................................................................... 73
Storage View Creation ....................................................................................................... 74
Storage Provisioning Methods ............................................................................... 78

© Copyright 2020 Dell Inc. Page iii
VPLEX Storage Provisioning Overview .............................................................................. 78
Advanced Provisioning Overview ....................................................................................... 79
Advanced Provisioning – Claim Storage ............................................................................ 80
Claiming Storage Volumes using VPLEX CLI .................................................................... 86
Advanced Provisioning - Create Extents ............................................................................ 87
Advanced Provisioning - Create Devices ........................................................................... 91
Advanced Provisioning - Create Virtual Volumes ............................................................... 96
Advanced Provisioning – Map Virtual Volume .................................................................... 99
EZ Provisioning Overview ................................................................................................ 102
EZ Provisioning ................................................................................................................ 103
VIAS Provisioning Overview ............................................................................................ 108
Register AMP and View Managed Arrays ........................................................................ 109
VIAS Storage Provisioning ............................................................................................... 110
VIAS Storage Provisioning ............................................................................................... 111
Provision-Job Rollback..................................................................................................... 115
Storage Volume Encapsulation ............................................................................ 116

Storage Volume Encapsulation Overview ........................................................................ 116
Using Storage Volume Encapsulation .............................................................................. 117
Encapsulation Method using CLI..................................................................................... 118
VPLEX Distributed Device Concepts.................................................................... 119

Distributed Devices Overview .......................................................................................... 119
Rule-Sets Handle Inter-Cluster Failures ........................................................................... 120
Distributed Device Detach Rule Options .......................................................................... 121
VPLEX Logging Volumes ................................................................................................. 122
VPLEX Consistency Groups ............................................................................................ 123
Consistency Groups: Visibility for Local Volumes ............................................................. 124
VPLEX Cluster Witness Benefit ....................................................................................... 125
Consistency Group Detach Rule Options ......................................................................... 126
Detach Rules - What if no Cluster Witness?..................................................................... 128
Continuous Availability with Cluster Witness .................................................................... 129
VPLEX Distributed Device Configuration ............................................................ 130

Page iv © Copyright 2020 Dell Inc.
Creating Distributed Devices using Dell EMC Unisphere ................................................. 130
Creating Distributed Devices from Existing Devices ......................................................... 131
Create a Distributed Device with the storage-tool compose Command ............................ 140
Create Distributed Devices using VPLEX CLI .................................................................. 141
VPLEX Consistency Group Creation ................................................................................ 149
Distributed Device Failure Scenarios ................................................................... 154

Distributed Device and Consistency Group Failure Properties ......................................... 154
Distributed Device Failure Scenarios ............................................................................... 156
Distributed Device Failure Scenarios ............................................................................... 159
Volume Expansion and Protection ....................................................................... 163

Virtual Volume Expansion ................................................................................................ 163
Procedure for Volume Expansion ..................................................................................... 165
Storage Volume Expansion Topologies ........................................................................... 166
Storage-volume Expansion Unisphere ............................................................................. 167
Storage-volume Expansion Procedure with CLI ............................................................... 168
Concatenation (RAID-C) Expansion Method .................................................................... 169
Concatenation Expansion – with CLI................................................................................ 170
Graphic View of Concatenation Expansion ...................................................................... 171
Volume Protection – Add a Local Mirror ........................................................................... 172
Volume Protection – Attach a Local Mirror with the CLI.................................................... 176
Adding a Remote Mirror ................................................................................................... 177
Data Protection with RecoverPoint ...................................................................... 182

VPLEX with RecoverPoint Protection ............................................................................... 182
Procedure for Adding RecoverPoint Protection ................................................................ 183
Importing RecoverPoint Certificate ................................................................................... 184
Create VPLEX Storage View for RecoverPoint ................................................................ 185
Adding RecoverPoint Cluster to VPLEX ........................................................................... 187
Create VPLEX Consistency Group .................................................................................. 188
VPLEX Data Mobility .............................................................................................. 189

Data Mobility Use Cases .................................................................................................. 189

© Copyright 2020 Dell Inc. Page v
Data Mobility Overview .................................................................................................... 190
Data Mobility Overview .................................................................................................... 191
Extent Mobility ................................................................................................................. 192
Device Mobility ................................................................................................................. 193
General Procedure to Perform Data Migration ................................................................. 195
VPLEX Data Migration Considerations............................................................................. 196
VPLEX Batch Migrations .................................................................................................. 197
Extent Mobility ................................................................................................................. 198
Monitor Extent Mobility Jobs ............................................................................................ 203
Device Mobility Wizard ..................................................................................................... 205
Complete Device Mobility................................................................................................. 210
Mobility Operations Through CLI...................................................................................... 211
Batched Mobility .............................................................................................................. 212
Create and Check the Batched Mobility Plan ................................................................... 213
Start and Cancel Batched Jobs ........................................................................................ 214
Pause and Resume Batched Jobs ................................................................................... 215
Monitor Batched Mobility Jobs ......................................................................................... 216
Commit, Clean and Remove Batch Jobs.......................................................................... 217
Role-Based-Access-Control .................................................................................. 218

Supported RBAC Roles ................................................................................................... 218
Accounts with “vplexuser” Role ........................................................................................ 219
Accounts with “readonly” Role.......................................................................................... 220
RBAC - View Account Role Information ........................................................................... 221
RBAC - Change Role for User ......................................................................................... 222
Shell Access Control - Enabled /Disabled ........................................................................ 223
If Restricted Shell Access - SCP Using "share" Folder ..................................................... 224
VPLEX Support Integration ................................................................................... 225

Configuring SRS Gateway ............................................................................................... 225
SNMP Overview............................................................................................................... 226
Supported SNMP Polling Commands .............................................................................. 227
Supported SNMP Trap Commands .................................................................................. 228
Using LDAPS for User Accounts ...................................................................................... 229

Page vi © Copyright 2020 Dell Inc.
LDAPS Configuration on VPLEX...................................................................................... 230
VPLEX Monitoring Concepts ................................................................................ 231

Monitor Clusters ............................................................................................................... 231
VPLEX Monitoring............................................................................................................ 232
VPLEX Performance Monitoring ........................................................................... 249

Overview of VPLEX Performance Monitoring ................................................................... 249
VPLEX CLI Performance Monitors ................................................................................... 250
Perpetual Performance Monitor Files ............................................................................... 251
Create Pre-configured Monitors ....................................................................................... 253
Verify Running Monitors ................................................................................................... 255
Manually Poll for Pre-configured Monitors ........................................................................ 256
Add Manual Polling to the Scheduled Tasks .................................................................... 257
Custom Monitor Configuration Steps................................................................................ 258
Determine the Type of Statistics to Collect ....................................................................... 259
Types of Statistics ............................................................................................................ 260
Example: How to Create a Custom Monitor ..................................................................... 261

© Copyright 2020 Dell Inc. Page vii
VPLEX Architecture
VPLEX Ecosystem
VPLEX is an important element in the Dell Technologies ecosystem. It is the core

of data protection when integrating with a wide range of storage arrays.
Storage
 Storage Arrays
o VPLEX is an important element in the EMC ecosystem and is the core of

data protection when integrating with storage platforms, such as VNX, Unity,
VMAX, XtremIO, and other third party storage.
 RecoverPoint
o RecoverPoint and RecoverPoint for VMs can provide continuous data

protection for VPLEX volumes.
 Applications
o Many applications, those that are supported on Block Storage, can run on
VPLEX volumes
Page
Internal Use - Confidential 8 © Copyright 2020 Dell Inc.
VPLEX Architecture
VPLEX at a Glance
Application Clusters stretched across data centers
• Continuous availability
• Data mobility without host disruption
70+ Non-Dell EMC Storage platforms
VPLEX Solution
Dell EMC VPLEX is an important part of the Data Protection and Availability
continuum. It delivers data availability and non-disruptive mobility across arrays in a
single data center or data centers separated by distance. VPLEX permits
technologies like VMware and other clusters that assume a single storage instance
and enabling them to function across arrays and across distance.
© Copyright
Internal Use - Confidential 2020 Dell Inc. Page 9
VPLEX Architecture
Availability Challenges
In a worst case scenario, what is the best possible

outcome IT could expect?
IT continues as if nothing happened.
Page
VPLEX Architecture
Data Mobility Challenges
Tech Refresh and Data Center Change Class of Maintenance Workload Balancing
Migrations Service
Traditional methods result in…
• Application downtime
• Poor IT resource utilization
• Migration service costs
• Months for Tech Refresh
• Risk of things going wrong
Over time, change to any infrastructure is unavoidable, especially in a cloud

environment and big data world. For example: with the advance and evolution of
technologies, data must be migrated to new storage arrays during a tech refresh. IT
resources may need to be reallocated for workload balancing purposes. Other
changes, like maintenance and changing the class of service (storage tiering) also
may be necessary.
© Copyright
VPLEX Architecture
Active-Active Data Center with VPLEX
ACTIVE Continuous data ACTIVE

Site availability, Site
1 2
deliver zero RPO/RTO
Stretched host clusters
VPLEX Cluster 1 VPLEX Cluster 2
Simultaneous R/W Non-disruptive data

migration across arrays
at both sites
or sites
VPLEX Metro
VPLEX storage virtualization and active-active technology leveraging VPLEX's

availability and mobility features. Host clusters and storage can stretch across
datacenters. The second data center is no longer sitting idle and can provide
continuous availability for mission-critical applications. VPLEX Metro configuration
allows VPLEX clusters at both data centers.
Page
VPLEX Architecture
Availability with VPLEX
The key component used to enable an active-active data center is VPLEX

Distributed Virtual Volumes.
Distributed Virtual Volumes have mirror legs at more than one cluster. Some of the benefits of
implementing a distributed active-active data center include:
 Increased availability - both data centers can serve production workloads while
providing high availability backup for the other data center.
 Increased asset utilization - passive data centers can have idle resources.
Active-active data centers make the most use of resources
 Increased performance/locality of data access - data need not be read from

the production site as the same data is read/write accessible at both sites
Site A Site B
Virtual Volume Virtual Volume
Distributed Virtual Volume
Synchronous
Latency
Up to 10 ms
© Copyright
VPLEX Architecture
VPLEX Product Family
VPLEX Metro
VPLEX Local Cluster 1 Cluster 2
The VPLEX product family is composed of VPLEX Local and VPLEX Metro
systems.
A VPLEX Local provides a seamless ability to manage and mirror data between multiple
heterogeneous arrays from a single interface. A VPLEX Local configuration consists of a single
VPLEX cluster. A VPLEX cluster is comprised of one, two, or four engines.
A VPLEX Metro enables active/active, block-level access to data between two sites within
synchronous distances. The distance is limited not only by physical distance but also by host
and application requirements. Depending on the application, VPLEX clusters can be installed
with inter-cluster links that have up to 10 ms round trip time (RTT). The combination of virtual
storage with VPLEX Metro and virtual servers enables the transparent movement of virtual
machines and storage across synchronous distances. This technology provides improved
utilization and availability across heterogeneous arrays and multiple sites.
Page
VPLEX Architecture
VPLEX for ALL Flash
 A VPLEX appliance model that packages VPLEX VS6 with Dell Technologies
all-flash systems
 XtremIO
 Unity All Flash
 Dell EMC PowerStore
 VMAX AF
 PowerMax
© Copyright
VPLEX Architecture
Where and When to Use VPLEX
Continuous Availability Data Mobility
Protect data in the Move data

VPLEX keeps event of disaster without IT needs to move
mission critical or failure disruption
data. VPLEX gets
applications
your data wherever
running even in
you want with
the face of
unplanned no planned
Collaborate over
disasters distance downtime
Zero RTO
Zero RPO
Always on, no matter what
Use VPLEX for:
 Protect data in the event of disasters or failure of components in your data

centers.
 Move data non-disruptively between Dell EMC and other third-party storage
arrays without any downtime.
 Collaborate over distance. Access Anywhere provides cache-consistent active-
active access to data across VPLEX clusters. Multiple users at different sites
can work on the same data while maintaining consistency of the dataset.
Page
VPLEX Architecture
VPLEX Use Case - Online Tech Refresh
No host disruption
New Array
Old Array
1. New array is Connected to the SAN

– Zoning and LUN Masking configured
2. VPLEX discovers a new array, and the admin creates migration target devices.
3. VPLEX Administrator sets up and starts a data migration job for each device.
4. The admin monitors the progress of the migrations. Hosts I/O's continue.
5. Once volumes on the new array are fully synchronized, the admin commits the
migration, which will disconnect from the mirror legs on the old array.
6. The old array can be removed without disruption.
© Copyright
VPLEX Architecture
VPLEX Use Case - Storage Pooling
 Pool arrays together

 Reduce cost without increasing risk
 Move devices 100% online between different tiers
 Mirror across arrays for enhanced availability
Virtual Volume
Physical
Device
TRANSPARENT MOBILITY
Page
VPLEX Architecture
VPLEX Continuous Availability
Continuous Availability = High Availability + Disaster Recovery
Stretched Host Clustering
Distributed Virtual Volume
Active Site Active Site
Here are some key Simple of VPLEX continuous availability:
 Simple DR testing
 Active assets in both sites
 No complex failover procedures
 Reduces planned and unplanned downtime
 Reduces TCO
© Copyright
VPLEX Architecture
Stretched VMware Clusters
≤ 10 ms *
Site A Site B
* Application specific
Site C
VPLEX Metro provides automatic load balancing between clusters.
A VPLEX Metro supports VMware HA and FT. It also provides VMware DRS and Vmotion
integration. For the VMs and applications to failover transparently, the data must be shared
across cluster nodes. VMware ESXi clustering requires shared storage to provide non-
disruptive movement of virtual machines. VPLEX Metro allows storage to span multiple data-
centers allowing ESXi servers in different failure domains to share access to Datastores
created on VPLEX Distributed Storage. VPLEX Metro fits perfectly with VPLEX distributed
cache coherence for automatic sharing and load balancing.
VPLEX supports VMware vSphere® Storage APIs - Array Integration (VAAI), also referred to
as hardware acceleration or hardware offload APIs . The APIs define a set of "storage
primitives" that enable the ESXi host to offload certain storage operations to the array, VPLEX,
which reduces resource overhead on the ESXi hosts and can significantly improve
performance for storage-intensive operations such as storage cloning, zeroing, and so on.
Page
VPLEX Architecture
VPLEX with RecoverPoint
WAN
Local Remote
Sync / Async
Protection Protection
IP or FC
Journal Journal
Local Copies Remote Copies
RecoverPoint with VPLEX
VPLEX and RecoverPoint products together offer continuous availability and operational and
disaster recovery. For customers requiring even further levels of availability, VPLEX can be
used with RecoverPoint to enable a third site for disaster recovery. This site can be located at
any supported distance, ensuring operational resilience in the event of a regional outage.
© Copyright
VPLEX Architecture
MetroPoint Solution
C
DR Site
Single DR Copy
B
A Disaster Recovery Production
Production
Site
Site Automatic Switchover
VPLEX Metro
Continuous Availability
CDP On Both Sides Of Metro
RecoverPoint RecoverPoint
Operational Operational
Recovery Recovery
Remote And Local Protection
VPLEX MetroPoint Configuration
VPLEX using RecoverPoint with a VPLEX Metro configuration allows for a unique configuration
referred to as MetroPoint topology. This MetroPoint topology provides a three or four site
solution. This allows for protection to continue if one of the VPLEX Metro Clusters fails.
Page
VPLEX Architecture
VPLEX Architecture
Dell EMC VPLEX Logical Architecture - Director I/O Stack
 VPLEX Front End
o Virtual Volumes are exposed to front-end hosts in storage views. Host

I/O and SCSI task management are done at this layer of the I/O
stack. The Front End of VPLEX functions like a storage array from the
host point of view. A VPLEX Storage View, which consists of FE
Ports, Host Initiators, and Virtual Volumes is the VPLEX method of
LUN Masking.
© Copyright
VPLEX Architecture
 Distributed Coherent Cache
o Virtual volume writes are cached and kept consistent and coherent
across all directors in a VPLEX system. Per-volume caching
implements local and global cache, and maintains consistency
between all directors in a VPLEX.
 Device Virtualization
o Device virtualization creates composite devices from storage

volumes. This allows for, creating local and distributed (VPLEX Metro)
mirroring. It also allows for Virtual Volume creation.
 VPLEX Back End
o The Back-end ports on VPLEX Directors function in the role of host

initiators to all storage arrays. The storage array ports are zoned to
VPLEX BE ports. Storage Volumes represent the devices masked to
the VPLEX BE ports from the back-end arrays. These Storage
Volumes are claimed by VPLEX and can be used to create Virtual
Volumes.
Page
VPLEX Architecture
VPLEX Constructs
Virtual Volume
Device
Extent
Storage Volume
Claim Storage
Storage
Array
LUN
Here is an explanation of the VPLEX Constructs:
 Virtual Volume A VPLEX Virtual Volume is created from the top-level Device. It is the storage
or Logical Unit presented to one or more hosts as part of a Storage View.
 Device A VPLEX Device is the application of a RAID topology to one or more extents. A Device
can be either a Local Device or a Distributed Device (VPLEX Metro topology is required). A
Device may use another device as an extent. Devices can be made up of any combination of
local devices and extents as appropriate for a particular RAID topology/geometry.
 Extent A VPLEX Extent is a slice or portion of the available space from a Storage Volume.
With VPLEX, you can create an extent that uses the entire capacity of the underlying Storage
Volume, or just a portion of the space. Extents provide a convenient means of allocating what is
needed while taking advantage of the dynamic thin allocation capabilities of the back-end array.
 Storage Volume Storage Volumes are the Logical Units (LU) presented from an array to the
VPLEX BE ports. These are initially unclaimed but can be claimed as part of the provisioning
process used by VPLEX. This can be done using the CLI of the VPLEX UI.
© Copyright
VPLEX Architecture
VPLEX Local Architecture
Here is an example of the different layers of a VPLEX Local Architecture:
Page
VPLEX Architecture
VPLEX Metro Architecture Builds on VPLEX Local
VPLEX Metro provides a unique solution to data center problems.
APPLICATION MOBILITY NOW POSSIBLE
STRETCHED HOST CLUSTER
Site 2
Site 1 Virtual Volume
IP or FC WAN
Displayed is a VPLEX Metro VMware HA solution. VPLEX Access Anywhere

technology allows you to export a single virtual volume from both VPLEX clusters
simultaneously.
In this solution, the physical hosts are only connected to their local VPLEX cluster
and no SAN extension between sites is necessary. This will help to isolate data
centers and prevent failures in one data center from affecting the other.
There are many interesting solutions that can be built atop VPLEX Metro. VMware
vMotion, VMware HA, and Oracle RAC are examples.
© Copyright
VPLEX Architecture
VPLEX Physical Architecture
VS2 Engine
Block Diagram VS2 Engine
Hosts
8 Gbps FC
8 Gbps FC
0 1 2 3 0 1 3
2
VPLEX VPLEX VPLEX
Front-end Front-end
Director FC Ports
Engine FC Ports Director
A B
Inter 1 1 Inter
Cluster Cluster
Distributed Cor Core COM Ports 0 0 COM Ports Cor Cor Distributed
Cache Cache
Cor Co Intra- 1 1 Cor Cor
Intra-
Cluster Cluster
COM Ports 0 0 COM Ports
8 Gbps FC
Back-end Back-end
FC Ports FC Ports
0 1 2 3 0 1 2 3
8 Gbps 8 Gbps FC
FC
Storage
Volume
VS2 Engine Details VPLEX architecture consists of the following components:
 Front-End (FE) ports are Fibre Channel ports which are zoned to host HBA
ports. They provide host connectivity to VPLEX Virtual Volumes
 CPU's and Distributed Cache They are key components for data
processing and storage virtualization.
 Back-End (BE) ports VPLEX BE ports allow Fibre Channel connectivity to

any supported storage array.
 Comm and WAN ports COM ports provide communication with other
directors in the local VPLEX cluster. Wan ports, both FC and IP, provide
communication with remote directors in a VPLEX Metro environment.
A VPLEX cluster can have up to four engines (or eight directors). Each director adds redundancy
and additional cache and processing power. VPLEX architecture is fully redundant to survive any
single point of failure. In any cluster, the fully redundant hardware can tolerate failure down to a
single director remaining with no Data Unavailability or Data Loss condition.
Page
VPLEX Architecture
VS6 Engine
Block Diagram VS6 Engine
Hosts
16 Gbps FC 16 Gbps FC
0 1 2 3 0 1 2 3
VPLEX VPLEX VPLEX
Front-end Front-end
Director Engine FC Ports
Director
FC Ports
A B
Inter- 1 1 Inter-
Co Co Cluster Cluste Co Co
Co Co Co COM 0 0 COM Ports Co Co Co
Co Co Co
Intra 1 1 Intra
Co Co Co Distributed
Distributed Co
Co Co Cluster Cluster Co Cache
Cache COM Ports 0 0 COM Ports
Back- 40 Gbps Back-

end FC IB end FC
0 1 2 3 0 1 2 3
16 Gbps FC 16 Gbps FC
Storage
Volume
VS6 Details
VS6 architectural design is the same as the VS2 architecture. There is still the
same number of FE ports, BE ports, and COM ports.
VS6 engines are much higher performing because they are built from higher-
performing hardware. VS6 directors get their performance boost by using dual six-
core processors, faster FE and BE ports, more cache, and InfiniBand for intra-
cluster communications.
© Copyright
VPLEX Architecture
External Comparison of VS2 and VS6 Engines
VS2 Technology
Director A Director B
SPS
Front View
Director B Director A
SPS
Rear View
Page
VPLEX Architecture
VS6 Technology
BB
Director B
Front View
Director A
BB
Director B
Rear View
Director A
© Copyright
VPLEX Architecture
Comparison
Compare VS2 and VS6 Hardware
VS2 VS6
CPUs per
director
2 six-core processors
1 quad-core processor 12 CPU cores
4 CPU cores
Memory per 36 GB (~24GB for Cache) 128 GB (~113 GB for Cache)

director
FE and BE 8 Gbps Fibre Channel 16 Gbps Fibre Channel

Connectivity
LOCAL COM Dual 8 Gbps FC Fabrics Dual 40 Gbps IB Fabrics

Connectivity
WAN COM Dual 8 Gbps FC or 10 GbE Dual 16 Gbps FC or 10 GbE

Connectivity
Page
VPLEX Architecture
Intra-Cluster Communications (Local COM)
VS2
VS2 Local COM
VPLEX Cluster
Directo Director
r 4A 4B
FC SW-B
Director Directo
3A r 3B
Director Directo
FC SW-B
2A r 2B
Director Director
1A 1B
VS2 uses
Fibre Channel (8 Gbps)for Local
COM connections
VS2 technology uses 8 Gbps Fibre Channel to communicate between directors.

These intra-cluster communication paths are used frequently when data is
requested at one director but resides in the cache of another director.
© Copyright
VPLEX Architecture
VS6
Local COM
VPLEX Cluster
Director Director
4A 4B
IB SW-B
Director Director
3A 3B
Director IB SW-B
Directo
2A r 2B
Director Director
1A 1B
VS6 uses
InfiniBand (40 Gbps)for Local COM
connections
VS6 engines not only have more cache per director, but the communication paths
(Local COM) between directors in the same cluster use InfiniBand. InfiniBand
provides 40 Gbps data paths between directors.
Page
VPLEX Architecture
VPLEX Local Rack Configuration
VPLEX Cluster Rack Configuration
Engine 4
SPS 4
Engine 3
Quad-Engine Cluster Cable Management

SPS 3
Engine 4
FC SW B
Cable Management & Laptop Tray

UPS B
FC SW A
Engine 3
UPS A
Cable Management
UPS B
Management Server
UPS A
IB Switch IB Switch InfiniBand Switch A and B

Engine 2
Cable Management
SPS 2
Engine 2
Cable Management
SPS 1
Cable Management
VS2 Cluster VS6 Cluster
VPLEX is configured as a single, dual, or quad engine cluster. A VPLEX Local will
have one cluster, and a Metro has two clusters connected together. A VPLEX Local
cluster consists of a single rack. Most installations use EMC factory-installed racks;
although VPLEX may be deployed in customer racks in the field. The rack contains
1, 2, or 4 VPLEX engines and each engine consists of 2 directors: A and B.Clusters
may be upgraded from one to two engines or upgraded from two to four engines
without disruption. Adding engines allows VPLEX to scale for greater performance
and redundancy.
© Copyright
VPLEX Architecture
VPLEX Engine I/O Modules
VS2
I/O Module details
A VS2 Engine contains two types of I/O modules, a 4 port Gbps and a 2-port 10
GbE. Here are where they can be used:
 Slot 0 is Fibre Channel and used for FE connections to a host.

 Slot 2 can be either 8 Gbps Fibre Channel or 10 GbE
o Used for WAN communications in a VPLEX Metro
 Slot 3 is 8 Gbps Fibre Channel used for local communication between
directors.
Page
VPLEX Architecture
VS6
I/O Module details
With VS6 technology, there are three types of I/O modules, a 4-port 16 Gbps Fibre
Channel module, a 10 GbE I/O module, and a 2-Port InfiniBand module.
Here are their functions and locations:
 Slot 0 has a 4-port 16 Gbps Fibre Channel module for FE connections to hosts.
 Slot 1 contains a 4-port 16 Gbps Fibre Channel module for BE connections to
storage arrays
 Slot 2 has either a 16 Gbps Fibre Channel module or a 10 GbE I/O module for WAN
COM (2 ports are used if a VPLEX Metro - otherwise ports are not used).
 Slot 2 is used for WAN communication between VPLEX
Clusters in a VPLEX Metro. 2 ports are used, in a VPLEX Local
the ports are not used.
 Slot 3 is a 2-Port InfiniBand module and is used for Local communications with
other directors in the same VPLEX Cluster.
© Copyright
VPLEX Architecture
WAN COM Performance Considerations
VPLEX Metro environment
FC vs. IP Round-trip- Bandwidth or Quality and

Time Capacity Reliability
VS6: 16 Gb/s FC
 Fibre Channel vs. IP - Choose a protocol for your VPLEX Metro WAN COM
connections. VPLEX Metro can be ordered with either 8 Gb/s Fibre Channel
(VS2) 16 Gb/s Fibre Channel (VS6), or 10 Gb/s Ethernet (VS2 and VS6) for the
WAN COM connections.
 Round-trip-time or WAN delay - The time to exchange data between clusters
will directly impact the distributed-device write time since the write mirroring is
synchronous. The maximum WAN round-trip-time (up to 10ms) is going to
largely be dependent upon what the applications can tolerate.
 Bandwidth or WAN link capacity - Ensure your inter-cluster WAN pipes are
sufficiently sized to handle the bandwidth required. Size appropriately for peak
capacity but also to tolerate a single link failure. Ideally VPLEX performance
should not suffer in the face of a single link failure, should one link happen to
fail, or be down for maintenance. Insufficient WAN COM bandwidth during times
of WAN saturation directly impacts latency and thus the host Metro write
performance. The amount of WAN bandwidth required depends primarily upon
the write rate for distributed-devices. Be aware of potential high bandwidth
users such as inter-cluster rebuilds, or reads in situations of a failed storage-
array.
 Quality and reliability - Details like packet loss dropped or corrupted frames or
lack of buffer credits can negatively impact the host Metro write latency.
Page
VPLEX Architecture
VPLEX Power Subsystem
Comparison of Power Subsystems
VS2 Technology VS6 Technology

PSU_B1 PSU_B0
Director A Director B
Director B
PS
0
PS
1
PS PS 1
Director A
VS2 Front PSU_A1

View VS
6 Rear PSU_A0
1100W Power Supply Units
Each Director has 2 Power Supplies for redundancy (N+1)
Independent power zones in the data center feed each VPLEX power zone,
providing redundant high availability.VS2 Technology: There are two power
supplies with fans per director. Both must be removed to pull the director out. To
ensure the power supply is completely inserted into the engine, the yellow at the
top of the power supply should not be visible.VS6 Technology: The Power Supply
Units (PSU) are N+1 technology. Each has enough wattage (1100 W) to run the
director and MMCS or MM (management module).
© Copyright
VPLEX Architecture
VPLEX Engine Battery Backup
VPLEX Power Backup System
VS2 Technology VS6 Technology

Dir. B - Battery Backup
Unit (BBU)
VS6 Front View

VS2 Rear View
SPS - Standby Power Supply

Dir. A - Battery
(Includes battery backup for both Directors A and B)
Backup Unit (BBU)
VS2 - Each engine is connected to 2 standby power supplies (SPS) that provide a
battery backup for cache vaulting in the event of transient site power failure. In
single-engine clusters, the management server draws its power directly from the
cabinet PDU.
In dual- and quad-engine clusters, the management server draws power from UPS-
A. Each VPLEX engine is supported by a pair of standby power supplies (SPS) that
provide a hold-up time of five minutes, allowing the system to ride through a
transient power loss. A single standby power supply provides enough power for the
attached engine.
Each standby power supply is a FRU and can be replaced with no disruption to the
services provided by the system. The recharge time for a standby power supply is
up to 5.5 hours. The batteries in the standby power supply can support two
sequential five-minute outages
VS6 - For VS6, there is no independent management server that needs to be

powered. The BBU (battery backup unit) pair provide 2 minutes of power to the
director and the MMCS or MM (management module). The dual BBU is not an
N+1 configuration, as both are required to provide the needed power.
Page
VPLEX Architecture
VPLEX Management Server
Management Server
VPLEX VS2 clusters have a separate management server that manages both A-
side and B-side directors. It has separate network connections for the customer
management network, and the A-Side and B-side internal subnets. It also has an IP
port for a service laptop connection. Either subnet can access any director. VPLEX
VS2 management server provides a management interface to the public network
for cluster management. It also has interfaces to other VPLEX components. All
VPLEX event logging, as a service, takes place on the management server.
© Copyright
VPLEX Architecture
VS6 Management Module Control Station
Management Module Control Station
VS6 does not have an external management server. Instead, VPLEX VS6
technology clusters have an integrated Management Module Control Station
(MMCS) that is part of engine 1 director A (MMCS-A). The MMCS also has all the
same network connections that the VS2 management server has.
Page
VPLEX Architecture
VPLEX VS6 MMCS IP Connections
IP Connections
Service Port (eth1)

128.221.252.2
253 Subnet
(eth0) B-Side
MRJ21 Cable Assembly

• .252 subnet A-side cable is lime-colored To Customer Management
Network (eth3)
• .253 subnet B-side cable is violet-colored
• Management network cable is black-colored 252 Subnet
(eth2) A-Side
The MMCS in engine 1 director A is used as the management server for the
cluster. There is a service port for connecting a service laptop (or the laptop that
comes with the rack) and an MRJ21 cable assembly that provides three more IP
ports:
- lime-green cable is for the .252 VPLEX internal subnet
- pink/violet cable is for the .253 VPLEX internal subnet
- black cable is used to connect to the data management network
© Copyright
VPLEX Architecture
VPLEX VS6 Management – MMCS
Management Control Station - MMCS
Only engine 1 has an MMCS in each director, and only the MMCS in director A
(MMCS-A) functions as a management server. The MMCS in engine 1 director B is
inactive for VPLEX CLI commands or the VPLEX Unisphere application.
An MMCS has its own CPU and 80 GB SSD which contains system code, logs,
and limited space for vaulting in case of power failure. There is no failover from one
MMCS to the other. If MMCS-A fails, it must be replaced.
The code on MMCS-B can be used to copy firmware to a new MMCS-A. MMCS-A-
Supports one Public IP for admin cluster management- Is used during EZ-Setup
and other VPLEX management operations.
It provides .252 subnet access to A-directors- Provides .253 subnet access to B-

directors.
Page
VPLEX Architecture
VPLEX VS6 Management - MM
MM - Management Module
 Installed in Engines 2,3,4

 One per Director
 Provides IP connectivity to internal VPLEX Management
o 252 subnet
o 253 subnet
© Copyright
VPLEX Architecture
VPLEX Management IP Infrastructure
VS2
Cable Connections
Engine 4
Cable Connections for Director and Switch Cable Connections for Director and Switch
management ports management ports
Engine 3
A-Side
B-Side
FC Switch B
FC Switch A
Management
LAN
Management Server
Management
Client
Engine 2
Engine 1
Shown are the Ethernet cables that connect the management ports from the VPLEX
management server to the management ports of each director and the Fibre Channel COM
switches. Note that there are No internal VPLEX IP switches and the directors are in fact daisy-
chained together on the “A” Side and the “B” side
The management server is the only VPLEX component that is configured with a public IP on
the data center management network. From the data center management network, the
management server can be accessed via SSH or HTTPS.
Page
VPLEX Architecture
VS6
Cable Connections
Engine 4
Cable Connections for Director
Cable Connections for Director management ports
management ports
B-Side A-Side
Engine 3
IB Switch IB Switch
Engine 2 Management
LAN
Management
Client
Engine 1
MMCS-A
The VS6 Management IP infrastructure is similar to the VS2. The major difference is that
instead of a separate management server, there is an embedded MMCS-A within engine 1
© Copyright
VPLEX IO Operations
VPLEX IO Operations
Distributed Cache Coherency
New Write: Rea

Block 3 Block 3
VPLEX’s distributed cache coherency handling enables superior availability. It enables any
director to service any I/O for any volume while participating in a global cache with all the other
directors in the cluster. Each director contributes cache to the global cache. If a director is
added to the cluster, the global cache increases in size. If a director fails and is removed from
the cluster, the global cache shrinks but access is maintained through all the remaining
directors. To illustrate how this works:
 When a write comes into director A for a particular block, a small piece of
metadata is updated to indicate that director A now has that block in cache.
 VPLEX does not provide acknowledgment to the host until data is stored on the
storage array.
 Should a read later come in for that same block on a different director, it will
look in the directory and see that the block is available in A’s cache. It will fetch
it from there and return it to the host.
Page
VPLEX IO Operations
VPLEX I/O Operations
Read Hit Distributed Device
HOST Stretched Host Cluster HOST
READ
GLOBAL CACHE DATA
VIRTUALIZATION
HETEROGENEOUS HETEROGENEOUS
FAILURE DOMAIN A FAILURE DOMAIN B
Host Read Request Read Hit:
 A host read request is serviced by a director.
 The director performs a lookup in its local cache.
 If the data is there it is sent to the host.
 If the data is not in the local cache a lookup is performed in the global cache.
 If another director, in that cluster only, has the data in its cache, then it is sent
via the local com to the servicing director and then to the host.
 The servicing director now has the requested data in its local cache, in order to
satisfy future potential reads.
© Copyright
VPLEX IO Operations
Read Miss Distributed Device

READ
GLOBAL CACHE DATA
VIRTUALIZATION
HETEROGENEOUS DATA
Read Miss
If the data the director is looking for is not in global cache, it is called a global cache miss. Here
is a description:
 A Read request is issued to a Virtual Volume
 A lookup is performed in local cache by the servicing director.
 A miss requires a lookup in Global Cache.
 On a miss from Global Cache, the requested data is read from the Storage
Volume into the local cache. The requested data is returned from the local
cache to the host.
 The servicing director now has the requested data in its local cache, in order to
satisfy future potential reads.
Page
VPLEX IO Operations
Write Local Device

DATA
GLOBAL CACHE DATA
VIRTUALIZATION
ACK
Write Local Device
When a Write is being committed to the array, the cache must first invalidate any existing
copies of that data in the global and local cache. During a write hit:
 A Write request is issued to the Virtual Volume.
 A lookup is performed in the local cache of the receiving director.
 Global cache updates the new location for the data.
 Any prior data is invalidated in all cache locations.
 The receiving director transfers the data to the local cache.
 The new data is written through to the back-end storage.
 The Write is acknowledged, first from the storage array, then VPLEX sends the
acknowledgement to the host.
© Copyright
VPLEX IO Operations
Write Distributed Device

DATA
DATA GLOBAL CACHE
VIRTUALIZATION
ACK
AC ACK
Write Distributed Device
Both VPLEX Clusters are Active-Active with a VPLEX Metro. Here are the steps for a Write to a
Distributed Device:
 Write is received by a director.
 The director identifies the blocks to be written and signals the other directors
that it now owns those blocks in cache.
o Both Local and Remote
 All directors update their private copy of the cache coherence table.
o Noting which blocks will now be invalid within their own caches.
 Both directors write the blocks to cache and through to Back-End storage.
 An acknowledgement is sent back from the "remote" cluster to the local cluster.
o At the same time the Local Cluster receives an acknowledgement from

the Back-End storage array.
 The Local Cluster acknowledges the host.
Page
VPLEX IO Operations
Path Redundancy
Single Path Failure
SAN SAN SAN SAN
FE Ports FE Ports
Director A Director B Director A Director B
Example:
Single Engine
VPLEX
Virtual Volume Virtual Volume
Virtual volumes presented from VPLEX to a host can tolerate path failures by
connecting the host to multiple directors and by utilizing multi-pathing software on
the host to control the paths. a virtual volume is presented out of multiple VPLEX
front-end ports on different directors. This yields continuous data availability in the
presence of port or director failure.
© Copyright
VPLEX IO Operations
Failure Across Engines
Engine Redundancy
SAN SAN SAN SAN
Engine 1 Engine 2 Engine 2
Example:
Dual Engine
VPLEX Virtual Volume Virtual Volume
Virtual volumes presented from VPLEX to a host can tolerate entire VPLEX engine failures by
connecting the host to VPLEX Front-End ports on different engines. An engine could fail and
the host would still be able to access its volumes. It is still best practice to connect the host to
one A director and one B director.
Page
VPLEX IO Operations
VPLEX Cluster Witness
Failure
VPLEX
Domain #3 Witnes
Is there a
failure?
Where should I/O
VP
continue?
N
Failure Domain #1 IP Management Failure Domain #2

VPN Network VPN
VPLEX Cluster-1 VPLEX Cluster-2
VPLEX Witness
VPLEX Witness can be deployed (best practice) at a third location to improve data availability
in the presence of cluster failures and inter-cluster communication loss. The VPLEX Witness is
implemented as a virtual machine in a separate failure domain. This eliminates the possibility of
a single fault affecting both a VPLEX Cluster and VPLEX Witness.
VPLEX Witness connects to both VPLEX clusters over the management IP network using VPN.
VPLEX Witness observes the state of the clusters and thus can distinguish between an outage
of the inter-cluster link and a cluster failure.
VPLEX Witness uses this information to guide the clusters to either resume or suspend I/O.
© Copyright
VPLEX Management Options
VPLEX Management Tools
Management Options
VPLEX HTML 5 Unisphere
Click to expand
A single VPLEX cluster is managed by logging into the management server on a VS2 system
or logging into MMCS-A on a VS6 system. In VPLEX Metro configurations, you can manage
both clusters from a single management connection.
The management server or MMCS-A coordinates data collection, VPLEX software upgrades,
configuration interfaces, diagnostics, event notifications, and some director-to-director
communication.
Both a VPLEX CLI and a GUI called EMC Unisphere for VPLEX are used to configure,
upgrade, manage, and monitor the VPLEX. The VPLEX CLI supports all VPLEX operations.
CLI commands are divided across a hierarchical context tree structure.
Page
VPLEX Management Security
Password Protected Accounts (role-based access control)
Role Description
service Used to configure VPLEX ( Service personnel only)
securityadmin Admin user - for user account management
vplexuser VPLEX management role used for provisioning, migrations etc. Cannot perform account
management
readonly Monitoring only
VPLEX accounts are password protected. Each account will be assigned a role. There are four
roles: service, securityadmin, vplexuser, and read-only. Most users will be assigned either
the vplexuser or readonly role. Only Dell Technologies service personnel (or partners
performing service) should use the account with the service role.
The securityadmin user should be restricted to administrators that manage VPLEX user
accounts. Additional security options include using an LDAPs server and creating a
Certification Authority (CA) on the VPLEX management server for the purposes of signing
management server certificates.
The VPlexcli command security create-ca-certificate creates a CA certificate file

and private key protected by a passphrase.
© Copyright
VPLEX CLI Context Tree (partial listing)
Command Information
Distribute Management
Clusters Data migration Engines Server Monitoring Notifications System default
d storage
notifications
bindings Ports
call-home
distributed-devices Et
snmp-traps
rule-sets Engine-<n-n> Eth1
directors Eth
director-<x-y> A Eth3 Directors
hardware director_A
ports Monitors
list of ports director_A_director
Device-migrations
firmware director_A_diskReportMonitor
list of devices director-<x-y-> B
director_A_portReportMonitor
Extent-migrations
hardware director_A_volumeReportMonitor
list of devices
ports director_B
list of ports
Monitors
firmware director_B_director
director_B_diskReportMonitor
fans
director_B_portReportMonitor
mgmt-modules
director_B_volumeReportMonitor
power-supplies
stand-by-power supplies
The VPLEX CLI is based on a tree structure like the structure of a Linux file system.
Fundamental to the VPLEX CLI is the notion of object context. The object context is determined
by the current location or pwd (Linux print working directory command) within the directory tree
of managed objects.
The object context is determined by the current location or pwd within the directory tree of
managed objects. The CLI is divided into command contexts. Some commands are accessible
from all contexts and are referred to as global commands. The remaining commands are
arranged in a hierarchical context tree. These commands can only be executed from the
appropriate location in the context tree.
Except for system-defaults, each of the sub-contexts contains one or more sub-context to
configure, manage, and display sub-components. Many VPLEX CLI operations can be
performed from the current context. However, some commands may require the user to change
to a different directory before running the command.
Page
Key Management CLI Root (/) Sub-contexts
Sub Functionality
Context
clusters/ Create and manage links between clusters, devices, extents, system volumes and virtual
volumes. Register initiator ports, export target ports, and storage
views
connectivity/ Configure connectivity between back-end storage arrays, front-end hosts, local directors,
port-groups and inter-cluster WANs
distributed- Create and manage distributed devices and rule sets

storage/
data-migrations/ Create, verify, start, pause, cancel, and resume data migrations of extents or devices
monitoring/ Create and manage performance monitors
recoverpoint/ Manage RecoverPoint options
The CLI is divided into command contexts. Command contexts contain commands
that can be accessed only from within that context. The commands under each
context are arranged in a hierarchical context tree. These commands can only be
executed from the appropriate location in the context tree. Understanding the
command context tree is critical to using the VPLEX command-line interface
effectively.
Except for the ‘system-defaults/’ sub-context, each of the sub-contexts contains

one or more sub-contexts to configure, manage, and display sub-components.
Command contexts have commands that can be executed only from that context.
The topmost context is the root context, or “/”. Although there are more, shown
here are the key root-level sub-contexts where commands can be accessed to
configure, manage, and monitor VPLEX clusters, storage, and host connectivity.
Some commands are accessible from all contexts. These are referred to as global
commands.
© Copyright
Navigating the CLI Context Tree
VPlexcli:/> cd /clusters
- Use cd command to navigate
VPlexcli:/clusters> cd cluster-1
- Find the current context from CLI interface prompt
VPlexcli:/clusters/cluster-1> cd connectivity
- Use ll command to displays the sub-contexts
VPlexcli:/clusters/cluster-1/connectivity> cd back-end
- Return to the root context with cd /

VPlexcli:/clusters/cluster-1/connectivity/back-end> cd port-groups
VPlexcli:/clusters/cluster-1/connectivity/back-end/port-groups> cd fc-port-group-0
VPlexcli:/clusters/cluster-1/connectivity/back-end/port-groups/fc-port-group-0> ll member-ports
/clusters/cluster-1/connectivity/back-end/port-groups/fc-port-group-0/member-ports:
Director Port Enabled Address

-------------- ------- ------- ------------------
director-1-1-A A1-FC00 enabled 0x50001442c00ef210
director-1-1-B B1-FC00 enabled 0x50001442d00ef210
VPlexcli:/clusters/cluster-1/connectivity/back-end/port-groups/fc-port-group-0>
Page
Locating in the Context Tree
Context Tree Details
The CLI includes several features to help locate your current position in the context
tree and determine what contexts and/or commands are accessible:
 The ls (list) command displays the sub-contexts immediately accessible from

the current context
 The ls -l (list long) command displays more information about the current
sub-contexts
 The cd command followed by a <Tab> displays the same information as ls at

the context level
 The tree command displays the immediate sub-contexts in the tree using the
current context as the root
 tree -e command displays immediate sub-contexts in the tree and any sub-
contexts under them.
© Copyright
Dell EMC Unisphere for VPLEX
Health Check
Unisphere for VPLEX provides many of the features that the VPLEX CLI provides, in a
graphical user interface (GUI) format. The GUI is very easy to navigate and requires no
knowledge of VPLEX CLI commands.
Operations are accomplished by clicking on VPLEX icons and selecting desired values. System
Status on the navigation bar shows a graphical representation of your system. It allows you to
quickly view the status of your system and some of its major components such as Directors,
Storage Arrays, and Storage Views.
System Status is the default screen when you log into the GUI. Also shown is the Monitoring
menu. Here, you can monitor VPLEX cluster performance, provisioning jobs status, and
general system health details.
Page
VPLEX Online Support
The Support page, located in the settings menu, provides links to various online
functions. These include VPLEX documentation, Help, and Solve Desktop.
© Copyright
VPLEX Connectivity
Back-End Array Zoning Best Practices - Unity
Zoning Best Practices
B B B2 B A0 A A2 A3
• Dual fabrics
VS2 / VS6 • Min 2 active paths / LUN
VPLEX
Engine
B A • Prefer 4 Active paths / LUN
• Distribute across engines
Back-End
Back-End
Fabric A Fabric B
Fabric-A Zones Fabric-B Zones
Zone Zone
E1_A1_FC00 E1_A1_FC01
Array_SPA_0 Array_SPA_0
Array_SPB_0
SP A SP B Array_SPB_0
Zone Zone2
E1_B1_FC00 E1_B1_FC01
Array_SPA_0 Array_SPA_0
Array_SPB_0 Array_SPB_0
Minimally Configured Back-End SAN
Ensure that you have a SAN implementation design that is consistent with the recommended
best practices. Consider each array allocating storage to hosts and their applications through
VPLEX. Here are a few best practice considerations when connecting VPLEX to back-end
arrays:
 Dual SAN fabrics should be used for redundancy.
 Each VPLEX director must have at least two active paths to every back-end
array storage volume presented to the cluster.
 Maximum of 4 active paths per Storage Volume per VPLEX director.
 Back-end connections should be distributed across multiple engines if possible.
Page
Back-End Zoning - Quad VPLEX with VMAX
Best Practices
This illustration shows the physical connectivity to a Dell EMC VMAX array. Similar
considerations should apply to other active/active arrays. Follow the array best practices for all
arrays including third party arrays.
 VPLEX initiators (backend ports) on a single director should spread across

engines to increase HA and redundancy.
 The VMAX Volumes should be provisioned for access through specific FA ports
and VPLEX ports.
 The VMAX Volumes within this grouping should restrict access to four specific
FA ports for each VPLEX Director ITL group.
© Copyright
Dell EMC XtremIO with VPLEX Configuration
VPLEX and XtremIO Details
XtremIO X-Brick Building Block With VPLEX, Balance load across

all target ports
The XtremIO Storage Array is an all-flash system, based on a scale-out architecture. XtremIO
storage array can include a single X-Brick, or a cluster of multiple X-Bricks. X-Brick cluster
scale from 2 to 16 active controllers simply by increasing the number of X-Bricks.
When connected to host through VPLEX, it is recommended to balance host access through
VPLEX between the X-Brick Storage Controllers to provide a distributed load across all target
ports.
Page
Determine WWNs of VPLEX FE Ports
Copy VPLEX FE WWPNs for Zoning to Host Initiators
VPlexcli:/> ls -l /engines/**/ports/ Port information output will be

/engines/engine-1-1/directors/director-1-1 A/hardware/ports: grouped by VPLEX engine and
director
Name Address Role Port Status

------- ------------------ --------- -----------
A0-FC00 0x50001442a0633000 front-end up
WWPNs
Determine the VPLEX front-end and WAN-COM port WWNs for use in configuring SAN
connectivity and zoning to support VPLEX-to-hosts and VPLEX cluster-to-cluster
communications. Use the VPLEX CLI ls –l command to list the contents of /engines/**/ports of
all VPLEX engines and directors.
© Copyright
Host Connectivity to VPLEX Front-End Ports
Host Connectivity Details
Zone Example
Fabric-B Zone_1
HBA-2
E1_A0_FC00
Fabric A SAN SAN E1_B0_FC00

Fabric B
F
C
F
C VPLEX FE
FC
FC Ports
Director B
02
FC
03
F
C
F
C
FC
01
Director A
FC
00
In a dual-SAN Fabric, best practice is to cross-connect each director’s FE ports into SAN
Fabric A and SAN Fabric B.
Page
FE Connectivity Best Practices
Best Practices:
Each host has four logical paths.

One path to Director-A on each fabric.
One path to Director-B on each fabric
Host based Multi-pathing software
installed.
Dual Fabric
VPLEX FE Ports per Director has a
minimum of two connections, one per
fabric.
Zoning
Provides redundant access to each Virtual

Volume from both A and B director in each
fabric.
© Copyright
VPLEX Front End Connectivity
Balance Between Directors
Fabric-A Zone_1 Fabric-B Zone_1

SAN A SAN B HBA-2
HBA-1
E1_A0_FC01 E1_A0_FC00
E2_B0_FC01 F
E2_B0_FC00
FC0
FC0
FC0
FC
FC VPLEX Engine-2
FC
F
FC
FC
FC
FC
FC0
FC
VPLEX Engine-1
FC
FC
Most host connectivity for hosts running load balancing software should follow the
recommendations for a dual-engine cluster. The hosts should be configured across two
engines and subsequent hosts should alternate between pairs of engines effectively load
balancing the I/O across all engines.
Zoning Example:
 Four Logical Paths

 Even Ports to Fabric B
 Odd Ports to Fabric A
 Each HBA Port to multi Engines
Page
VPLEX Storage Provisioning Concepts
Storage View: Establish Host-to-Virtual-Storage Connectivity
Storage View Details
Storage View
Registered Host Initiators
VPLEX FE Ports
VIRTUAL
VOLUMES
Host
A Storage View is a combination of registered initiators, VPLEX Front-End ports, and Virtual
Volumes. It is used to control a single or clustered host access VPLEX Virtual Volumes. It is the
VPLEX method of LUN Masking.
To export VPLEX storage, you must first create a storage view for the host. Next add VPLEX
front-end ports and VPLEX Virtual Volumes to the view. Virtual volumes are not visible to hosts
until they are in a storage view with associated ports and initiators.
A registered initiator can be in more than one storage view and a VPLEX FE port can be in
more than one storage view, while the unique/particular combination of a specific
<initiator><FE_port> pair can only be in one storage view.
© Copyright
VPLEX Virtual Storage Constructs – Complex Device

Example
Complex Device Details
Registered Host Initiators
HOST
Top-Level Device
Back-end storage arrays are configured to present LUNs to VPLEX through the SAN. Each
presented back-end LUN maps to one VPLEX storage volume. Storage volumes are initially in the
‘unclaimed’ state. Unclaimed storage volumes may not be used for any purpose within VPLEX
other than to create meta volumes that are for system internal use only.
Once a storage volume has been claimed within VPLEX, it may be split into one or more contiguous
extents. A single extent may map to an entire storage volume. However, it cannot span multiple
storage volumes. A VPLEX device is the entity that enables RAID implementation across one or
more extents or other devices. VPLEX supports RAID-0, RAID-1, RAID-C, as well as 1-1 mapping.
Raid-0 can stripe data across multiple extent/device constructs. When creating a Raid-0 device, if
more than one extent is chosen, VPLEX creates a Raid-0 device that is striped across the selected
extents. The Raid-0 device is the sum of the size of the extents. An example would be if three 2 GB
extents were selected, then the Raid-0 device would be 6 GB. VPLEX would stripe data across the
selected extents. The stripe depth specifies how much data is written to an extent before moving to
the next extent.
Raid-C concatenates (appends) extents to provide a larger address space.
Raid-1 mirrors two device extent/device constructs. The top-level device is the same size. A storage
view is the masking construct that controls how one or more VPLEX virtual volumes are exposed
through VPLEX front-end ports to host initiators.
Once a storage view is properly configured and operational, the host should be able to detect and
use virtual volumes. A Host discovers virtual volumes presented by VPLEX, after initiating a bus-
scan on its HBAs. Every front-end path to a virtual volume is an active path and the current version
of VPLEX presents each virtual volume. The host requires multi-pathing software for a high-
availability implementation.
Page
View Virtual Volume ID Presented to Host
export storage view summary
VPlexcli:/clusters/cluster-1/exports/storage-views> export storage-view summary
View health summary(cluster-1):

view name | health-state exported volumes ports registered initiators
---------------------------------- ---------------- ------------------------ ------- --------------------------
Student5_StorageView healthy 4 4 2
Total 1 views, 0 unhealthy.
VPlexcli:/clusters/cluster-1/exports/storage-views> export storage-view map -v *
VPD83T3:6000144000000010e00c2ecb3c5914fb Exchange__1_vol
VPD83T3:6000144000000010e00c2ecb3c59152a Dev_14_vol
VPD83T3:6000144000000010e00c2ecb3c591530 Dev_8_vol
VPD83T3:6000144000000010e00c2ecb3c591549 Dev_10_vol
ID of Virtual Volume Virtual Volume

presented to host name
Use the export storage-view summary command from the context shown to see a
summary of storage views that are configured in VPLEX.
A unique VPD (vital product data) ID is assigned to each VPLEX virtual volume. We can view
this ID by entering the VPLEX CLI command export storage-view map
<storage_view>. This is the same logical device ID that is seen in all host operating systems
to identify a LUN from a Storage System. This VPD number will not change even if the
underlying storage is moved.
Here we see an example of a PowerPath CLI command. Notice the logical device ID is the
same as the VPD of a VPLEX Virtual Volume.
© Copyright
Storage View Creation
Register Initiators
The first step when creating a Storage View is to discover and verify the
connectivity of the VPLEX Front End ports.
Initiators can be given a text name.
Select Ports
VPLEX FE Ports are

added to the
Storage View. The
ports selected must
be zoned to the HBA
ports.
Page
Select FE Ports
© Copyright
Add Virtual Volumes
Volumes here will be exported to host
Page
Review and Complete
Click Finish to create the Storage View.
© Copyright
Storage Provisioning Methods
VPLEX Storage Provisioning Overview
To begin using VPLEX, you must provision storage so that hosts can access that
storage.
There are three ways to provision storage on VPLEX:
 Advanced provisioning
 EZ provisioning
 Integrated array service-based provisioning (VIAS)
Page
Advanced Provisioning Overview
Advanced Provisioning Details
View and
claim Create
Create RAID-
available Extents
0, RAID-1,
storage from
RAID-C or
Storage Create Place
1:1 mapping
Volumes Virtual Virtual
of Extents to
Volumes Volumes
Devices
into a
Storage
View
© Copyright
Advanced Provisioning – Claim Storage
Claim Storage
Select Cluster and Storage Volumes
Select unclaimed
Storage Volumes
To claim a Storage Volume, select:
 Provision Storage
 Desired Cluster and Storage Volume view
 Unclaimed Storage Volumes
Page
Launch the Wizard
Launch the Wizard
 Use the "More" drop-down menu

 Select Claim All Storage on Supporting Array
© Copyright
Step One
Select the desired array from the list.
Page
Step Two
Mapping File Selection (Optional)
© Copyright
Step Three and Four
Storage Volume selection can be altered here. The arrows move Default Names can be edited.
volumes into the right column.
Page
Storage Volume Claimed
A view of the Storage Volumes will display all that have been claimed.
© Copyright
Claiming Storage Volumes using VPLEX CLI
Storage-volume claimingwizard Command
The CLI command claimingwizard finds unclaimed storage volumes, claims them, and
names them appropriately. This command can be used to claim and name many Storage
Volumes with a single command.
Storage volumes must be claimed, and optionally named before they can be used in a VPLEX
cluster. Storage tiers allow the administrator to manage arrays based on price, performance,
capacity, and other attributes. If a tier ID is assigned, the storage with a specified tier ID can be
managed as a single unit. Storage Volumes without a tier assignment are assigned a value of
‘no tier’.
Optional arguments
Page
Advanced Provisioning - Create Extents
Extent Creation Wizard
We will now create the next level of volume construct, an extent. Select the desired Cluster,
then change the view to Storage Volumes. Use the drop-down menu to select Create Extents.
© Copyright
Select Storage Volumes
Select Claimed Storage Volumes
Storage Volumes selected automatically appear. This can be altered using the arrows. Only
claimed Storage Volumes will be displayed.
Page
Review
Selected Storage Volumes can be verified before Extents are created.
© Copyright
Verify
The results of Extent creation are displayed. They can be seen by changing the
"View By" to Extents for the specific VPLEX Cluster.
Page
Advanced Provisioning - Create Devices
Launch Device Creation Wizard
Select Extents for the View By window. Click Create Devices to launch the Wizard.
Select Extents
© Copyright
Step One
Device Type
Here, we can select a RAID protection or performance attribute or simply specify a one-to-one
mapping of one extent used to create a single device.
Page
Step Two
Select the extent to be used for the new device. Since we previously selected Raid-1, we must
select two extents minimum. Click the Add Device button to create the device.Data will be
copied from Source to Target.
Extent Selection The new Device can now be named.
Determined by previous selection
Raid-1 Select source and target
© Copyright
Step Three and Four
Virtual Volumes can be created here. If multiple Devices are created, a base name can be
given. Click Finish to complete the wizard.
Click to Expand
Page
View Created Devices
Devices can be viewed by Cluster.
New Device
Mirror Leg Syncing
© Copyright
Advanced Provisioning - Create Virtual Volumes
Create Virtual Volumes
Virtual Volume Creation
Select Devices Create Thin Virtual Volumes
Select Devices
Page
When creating Virtual Volumes, there is an option to make them thin enabled.
To make a Virtual Volume thin enabled several characteristics must be true:
 Storage volumes are provisioned from storage arrays that are supported by VPLEX as thin-capable
 All the mirrors are created from the same storage-array family that VPLEX supports (For a RAID-1 configuration)
 Storage volume display thin properties
This allows us to use host-based storage reclamation using the unmap feature of VMware ESXi hosts. For example, after
deleting a VM from a datastore, it is desired to reclaim the storage to use for other VMs. VMware VAAI (vStorage API Array
Integration) supports this feature.
If the virtual volumes cannot be created as thin, the operation will succeed but it will be thick instead.
Refer to the Dell EMC VPLEX GeoSynchrony Administration Guide for additional information.
© Copyright
Virtual Volume Wizard
Virtual Volume Created Virtual Volumes Displayed
Not in a Storage View
Page
Advanced Provisioning – Map Virtual Volume
Virtual Volume Mapping
To add Virtual Volumes to an existing Storage View, select the VPLEX Cluster, and change the
view to Storage Views. Click Add to begin.
© Copyright
Storage View Editing
Storage View Wizard Complete
Page
Verify Storage View
Storage View Properties
Verify the Virtual Volumes have been added to the Storage View. Change the View By to
Storage Views. Select Virtual Volumes in the Storage View Properties window.
© Copyright
EZ Provisioning Overview
EZ provisioning is a simple method of provisioning that makes provisioning tasks easy.
EZ provisioning automatically creates a virtual volume with a one-to-one mapping to a

selected unclaimed storage volume. Use EZ provisioning to quickly create a virtual volume
that uses the entire capacity of the storage volume. In EZ provisioning, you select storage
arrays and define how you want them to be used, protected, and presented to hosts.
Steps for EZ Provisioning:
1. Select or create a consistency group for the volumes

2. Select the type of volume to create (distributed or local).
3. Select protection options, storage, and optionally expose the virtual volume to
hosts.
4. Add hosts
Page
EZ Provisioning
Consistency Group
The first step is to select an existing Consistency Group or create a new one.
VPLEX Consistency Groups aggregate volumes to enable the application of a
common set of properties to the entire group. Consistency Groups are explained in
detail later in the course.
© Copyright
Volume Options
Volume Option Details
Descriptive Name
Select the source VPLEX cluster that will provide the storage capacity from the back-end
arrays. Also, select the appropriate protection and data synchronization attributes for the new
storage capacity.
Page
Storage Volumes
Storage Volume Selection Thin Attribute
Select a back-end storage array and LUN connected to the source VPLEX cluster. The LUN
data of the selected physical array will be copied onto the Storage Volume in the selected
VPLEX cluster. Back-end array LUNs in either the claimed or unclaimed state may be used.
VPLEX does not report a volume as thin to host initiators until its thin-enabled option is set to
true. This can be set here.
© Copyright
Final Steps
Add the new Virtual Volume to a Review the selections. Results

Storage View or create it as
unexported.
Page
Storage View
View the new Virtual Volume.
© Copyright
VIAS Provisioning Overview
VIAS details
Registered Storage
Initiators Views
VPLEX FE
Ports
Admin running Virtual
Unisphere for VPLEX Volumes
Storage Pool
VPLEX Management Server
AMP
AMP
• VIAS = VPLEX Integrated Array Services
• AMP = Array Management Provider XtremIO VMAX VNX
The VPLEX Integrated Array Services (VIAS) feature enables VPLEX to provision storage for
Dell EMC VMAX, VNX, and XtremIO storage arrays directly from the VPLEX CLI, UI, and
REST API. VPLEX uses Array Management Providers (AMPs) to streamline provisioning and
allows you to provision a VPLEX Virtual Volume from a pool on the storage array.
The VIAS feature uses the Storage Management Initiative-Specification (SMI-S) provider to
communicate with the arrays that support integrated services to enable provisioning. The SMI-
S provider is used for VMAX and VNX.
After the SMI-S provider is configured, you can register the SMI-S provider with VPLEX as the
Array Management Provider (AMP). When the registration is complete, the managed arrays,
pools, and storage groups are visible in VPLEX, and you can provision Virtual Volumes from
those pools. The pools used for provisioning must have been previously created on the storage
array, as VIAS does not create the pools for provisioning.
VIAS also supports a REST AMP used with XtremIO arrays. The REST AMP does not require
additional software. The provider is on the XtremIO array itself.
You need to register the AMP as type REST within VPLEX.
Each XtremIO array is registered for VIAS in a 1-to-1 relationship with a VPLEX cluster.
Multiple XtremIO arrays need to be individually registered in VPLEX. This is different from SMI-
S AMPs where multiple storage arrays are managed by one SMI-S provider, then the SMI-S
provider is registered with VPLEX.
Page
Register AMP and View Managed Arrays
Before provisioning storage using VIAS, we must register the Array Management
Provider (AMP).
Once the AMP is registered, we can see which arrays it manages and the free
space on the storage pools for each array.
SMI-S for all

other Arrays
Here is an explanation of the fields.
 Provider Type- Rest for XtremIO and SMI-S for AMPs

 Name - identify the XtremIO or AMP host
 IP Address - Address of the XtremIO or SMI-S host
 User/Password - Determined by the SMI-S host or XtremIO
© Copyright
VIAS Storage Provisioning
Using the VPLEX Integrated Array Services (VIAS) feature, you can create virtual
volumes from pre-defined storage pools.
To launch the VIAS wizard:
 Select the cluster

 View by Virtual Volumes
 Select Create>Provision from Pools
Page
VIAS Storage Provisioning
Here are the steps to configure a Virtual Volume using the Provision from Pools
Wizard.
Steps One and Two
Select the Consistency Group within the appropriate VPLEX Cluster(s) to use or
create a new one.
Consistency Group Volume Options
© Copyright
Steps Three
Storage Pools -This step selects the Back- Create Thin Virtual Volumes
End Storage. The array selection list is
based on the arrays added to the SMI-S
server, or the number of XtremIO arrays
added.
Page
Steps Four and Five
Storage Views The Virtual Volumes can be Review the selections made.
optionally added to a Storage View. The
Storage View must already exist.
© Copyright
Step Six
The final results are displayed. Provisioning may take View the status from the Jobs view.
longer.
Page
Provision-Job Rollback
VIAS Rollback
Undo operation or rollback steps are added in case the VIAS provisioning fails. Not
all steps in the VIAS will be rolled back. Here are the main create and rollback
steps. The pre-check does not need a rollback if it fails. The volume creation is
rolled back in the VIAS process, as well as the volume exposure to VPLEX. Steps
2 and 3 create 90% of the possible issues. Steps 4 and 5 do not have rollback
steps. Note: if an error occurs in step 4 or 5, provisioning artifacts will remain and
need to be deleted by the user.
© Copyright
Storage Volume Encapsulation
Storage Volume Encapsulation Overview
Database Server
Log Boot
Data
SAN
Log
Each Storage
Data Volume to be
Encapsulated
Management
Boot
Station
 Encapsulation claims Storage Volumes while retaining all existing data.

 Use encapsulation to import existing non-VPLEX volumes into VPLEX.
 Feature can be implemented using VPLEX CLI or Unisphere.
 CLI allows for scripting
Page
Using Storage Volume Encapsulation
Write down the

Un-mount the
UID of the
back-end
existing back-end
Storage Claim and
array LUN
Volume from Encapsulate the
Provision the new
host then Storage Volume
Virtual Volume back
detach using the VPLEX
to the host and note Remount the new
CLI
the new UID under Virtual Volume to
VPLEX the host
Steps for Storage Volume Encapsulation - VPLEX provides the ability to claim back-
end Storage Volumes already in use under its control. The process of claiming a storage
volume while saving existing user data is called storage encapsulation. Any existing back-end
user volume may be encapsulated, including non-bootable and host boot image volumes.
© Copyright
Encapsulation Method using CLI
Use the Storage-Tool Compose CLI command for Encapsulation.
storage-tool compose -n new_Encap -g raid-0 -d VPD83T3:60000970000195901096533030313832
-v /clusters/cluster-1/exports/storage-views/esx_21
Argument Definitions:
-n new_Encap specifies the name of virtual volume
-g raid-0 specifies the desired geometry to use -d [storage-volumes] - specifies a

list of storage volumes to be used to build the Virtual Volume.
-v [Storage View name] specifies the VPLEX storage-view(s) to receive the new
virtual-volume. In this example, the Encapsulated Volume is being exported to the
host esx_21 Storage View.
Page
VPLEX Distributed Device Concepts
Distributed Devices Overview
Mirr Mirr
or or
Distributed Devices - VPLEX Metro storage objects having a RAID-1 geometry with a mirror-
leg in each VPLEX Cluster.
 Distributed Devices support Virtual Volumes, that are presented to hosts

through a Storage View on each cluster.
 Writes issued to a Distributed Device are synchronously mirrored to both
clusters.
 Distributed coherent cache preserves data integrity.
© Copyright
Rule-Sets Handle Inter-Cluster Failures
Mirror Leg Mirror Leg
Rule-sets are predefined rules that determine which cluster continues servicing I/O when
connectivity between clusters is lost. If the VPLEX Metro clusters lose contact with one
another, or if one cluster fails, Rule-sets define which cluster continues operation. This cluster
is referred to as the "preferred cluster". The remaining cluster, the non-preferred, suspends I/O.
 Determine which cluster continues I/O after a failure.

 Inter-cluster link fails
 VPLEX Cluster fails
 Failure starts delay timer (5 seconds default)
Page
Distributed Device Detach Rule Options
Detach Rule Cluster-1 Cluster-2
cluster-1-detaches (Cluster-1 is Services I/O Suspends I/O

preferred)
cluster-2-detaches (Cluster-2 is Suspends I/O Services I/O

preferred)
 Distributed Device Rule-set may be overridden by Consistency Group detach

rules.
 Auto-resume setting determines behavior after connectivity is restored
© Copyright
VPLEX Logging Volumes
Logging Volume Logging Volume
Logging Volume details - Logging volumes are required at each cluster before a Distributed
Device can be created. Logging volumes are used to keep track of blocks written during an
inter-cluster link failure, or when one leg of a distributed RAID-1 becomes unreachable and
then recovers.
After a WAN link failure is restored or an unreachable leg recovers, VPLEX uses the
information in logging volumes to synchronize the mirrors by sending only changed blocks
across the link. After the inter-cluster link or leg is restored, the VPLEX system uses the
information in the logging volumes to synchronize the mirrors by sending only changed blocks
across the link. Logging volumes also track changes during the loss of a volume when that
volume is one mirror in a Distributed Device.
During and after link outages, logging volumes are subject to high levels of I/O. Thus, logging
volumes must be able to service I/O quickly and efficiently. For more information about logging
volume requirements and configuration, please see the VPLEX Administration Guide.
 Keeps track of write I/Os during an inter-cluster link outage or loss of access for
a Distributed Device.
 Required at each VPLEX Cluster before creating a Distributed Device.
 Log information used to synchronize mirrors after access is restored.
Page
VPLEX Consistency Groups
Consistency Group Details - VPLEX Consistency Groups aggregate volumes to enable the
application of a common set of properties to the entire group. Consistency Groups ensure the
same winning cluster for all the Virtual Volumes within the group during an inter-cluster
communication failure. Consistency group detach rules define on which cluster I/O continues
during cluster or inter-cluster link failures. The groups work together with VPLEX Cluster
Witness. The properties of a consistency group are applied to all the virtual volumes in the
consistency group. Here is a summarized list of the properties that can be applied:
 Cache mode
 Visibility
 Storage at cluster
 Local read override
 Detach Rule
 Auto resume at loser
 RecoverPoint enabled
Application A Application B
Consistency Group Consistency Group
Virtual Virtual Virtual Virtual Virtual Virtual Virtual

Volum Volum Volum Volum Volum Volum Volum
Cluster-1 Cluster-2
Consistency Groups ensure the same winning cluster for all the Virtual Volumes
within the group during an inter-cluster communication failure. Common properties
applied to all Virtual Volumes in the group. Works with VPLEX Cluster Witness
© Copyright
Consistency Groups: Visibility for Local Volumes
Visibility Details
VPLEX Cluster-2
VPLEX Cluster-1
…with Global Visibility
Local Consistency Group… W A
Virtual Volume
W = Write
A = Access (read)
Visibility controls which clusters know about a Consistency Group. By default, the visibility
property of a Consistency Group is set only to the cluster where the group was created. This is
referred to as local visibility. This means only hosts attached to the local cluster have
read/write access to the volumes in the consistency group. For global visibility, set the visibility
to both cluster-1 and cluster-2. With global visibility, host on both clusters have read/write
access to the volumes in the consistency group.
The visibility of the volumes within the consistency group must match the visibility of the
consistency group. Local Consistency Groups with global visibility will always be synchronous.
Page
VPLEX Cluster Witness Benefit
Maintains Distributed Consistency Group Availability - VPLEX Cluster Witness is an

optional component installed as a virtual machine. Cluster Witness helps VPLEX Metro
configurations automate the response to cluster failures and inter-cluster link outages.
IP Management
Network
Inter-cluster
Network A
Inter-cluster
Network B
Witness Benefits:
 Automatically distinguishes between and provides guidance to clusters on how

to react to a cluster or inter-cluster communication failure.
 Allows VPLEX Metro to provide continuous availability in case of a failure.
 VPLEX Cluster Witness VM must be in a third failure domain relative to the
VPLEX cluster sites to be effective.
 Works only in conjunction with distributed consistency groups.
© Copyright
Consistency Group Detach Rule Options
Detach Rule Option Details
Detach Action
Rule
winner cluster-  Cluster specified by cluster-name is declared the preferred cluster

name delay
seconds
if a failure lasts more than the number of seconds specified
 Detach rules apply to all volumes of a consistency group and
override rules applied to individual volumes
 Preferred cluster may be guided by VPLEX Cluster Witness (if
available)
no-automatic-  NOT guided by VPLEX Cluster Witness

winner
 Detach rules of the member devices determine the preferred
cluster for that device
Page
The Consistency Group Detach Rule designates which cluster detaches if clusters lose
connectivity. Possible values:
 No-Automatic-Winner - the consistency group does not select a winning

cluster.
 Winner (cluster-name) (delay) - The cluster specified by cluster-name is

declared the winner after the inter-cluster link outage lasts more than the
number of seconds specified by delay. For a Consistency Group that is
participating in RecoverPoint, this value must be set to the cluster at which the
RecoverPoint Splitter is running.
If a consistency group has a detach-rule configured, the rule applies to all volumes in the
Consistency Group and overrides any rule-sets applied to individual volumes.
© Copyright
Detach Rules - What if no Cluster Witness?
Detach Rules Details
I
Failure of Cluster-1 I
I I
I I
I
I
I
I Cluster-2 loses communication with
I Cluster-1.
I
I I
By rule, Cluster-2 suspends I/O
I I
Non-preferred
I I
O * Data Unavailable *
Winner
I I
(static bias)
I
Stora
ge
Stora View
ge
Cluster-2
Cluster-1
Distributed Raid-1 Device
Mir Mirr
Device ror or Device
Rule: “Cluster-1 Detaches” Leg
For a VPLEX Metro without Cluster Witness, there has to be a method to avoid split-brain
scenarios. Each Distributed Device has a Rule-set applied to it.
As discussed, Rule-sets are predefined rules that determine which cluster continues I/O when
connectivity between clusters is lost. When a loss of connectivity occurs, VPLEX starts a delay
timer (default 5 seconds) and suspends I/O to all Distributed Devices on both clusters. If
connectivity is not restored before the timer expires, then the rule is enforced.
Without VPLEX Cluster Witness, we may have a scenario as shown. The Distributed Device
has the detach Rule set to "Cluster-1 detaches". If cluster-1 fails, then cluster-2 uses the rule
and suspends I/O. Now data is unavailable until the problem is resolved.
Page
Continuous Availability with Cluster Witness
Continuous Availability Details
C C
C
Cluster Witness
Cluster-1 Cluster-2 Cluster-1 Cluster- Cluster-1 Cluster-2 Failures
2
Cluster-1 and 2 both continue I/O; Cluster-1 and 2 will continue I/O; Cluster-1 and 2 will continue I/O;
Call home is issued Cluster-1 issues the call home Cluster-2 issues the call home
CW CW C
Cluste
Cluster-1 Cluster-2 Cluster- Cluster-
Cluster-2 Cluster-1 r
1 2
Cluster with preferred status,

Cluster-2 fails; Cluster-1 fails;
will continue I/O;
CW Guides Cluster-2 to continue I/O
Other cluster will suspend I/O CW Guides Cluster-1 to continue I/O
No Data Unavailability
VPLEX Cluster Witness helps VPLEX Metro systems with consistency groups respond to
cluster failures and inter-cluster (WAN COM) link outages. Presented here are the various
inter-site failure scenarios and how they are handled when using VPLEX Cluster Witness.
 If the Cluster Witness host fails, each cluster will lose communication with
Cluster Witness and call home. I/O continues normally.
 If management link between cluster and CW hosts fails, then the cluster that
detects this will call home.
 If the WAN COM link between clusters fails, each cluster suspends I/O until
receiving guidance from Cluster Witness. Cluster Witness gives guidance to
each cluster to default to the rule set (where preferred cluster continues I/O and
the non-preferred cluster suspends I/O).
 If Cluster-1 fails, Cluster-2 marks its cluster-witness operational-state as

Unknown upon seeing the failure. It takes about 4 seconds after that to receive
the guidance decision from Cluster Witness Server. Cluster Witness guides
Cluster-2 to continue I/O.
 If Cluster-2 fails, Cluster-1 marks its cluster-witness operational-state as

Unknown upon seeing the failure. It takes about 4 seconds after that to receive
the guidance decision from Cluster Witness Server. Cluster Witness guides
Cluster-1 to continue I/O.
© Copyright
VPLEX Distributed Device Configuration
Creating Distributed Devices using Dell EMC Unisphere

Distributed Devices are mirrored between clusters in a VPLEX Metro. They support virtual
volumes, that are presented to a host through a storage view. All Distributed Devices must be
associated with a Logging Volume. During a link outage, the logging volume is used to map the
differences between the mirrored legs. All Distributed Devices must have a detach Rule-set to
determine which cluster continues I/O when connectivity between clusters is lost. Using the
wizard requires adding the devices to a Consistency Group. A Consistency Group will have it's
own Rule-set configured. This will override the device Rule-set.
To create Distributed Devices from Storage Volumes, navigate to Provision

Storage>Distributed Storage>Distributed Devices> and choose the Distributed Device type.
Select Create. This launches the wizard.
Creating Distributed Devices from Storage Volumes
Page
Creating Distributed Devices from Existing Devices
Device Selection
Verify before launching the Wizard which existing Devices on each cluster are to be
used.
Considerations for selection:
 Source and target selection

– Data is synchronized from source to target.
 Device size
– Target Device must be equal to or greater than the Source device
 Devices may not have Virtual Volumes created.
© Copyright
Launch the Wizard
Select Provision Storage and View By Devices.
Page
Select Devices
Select the source cluster, then choose the Device.
© Copyright
Select Mirror
The next step is to select the target device. possible candidates are based on the
size of the source Device.
Click Add Mirror after

Devices are selected
Page
Synchronize
Best practice is to synchronize.
© Copyright
Consistency Group
Create a new group, or select an existing one.
Page
Create Device
Review selections
© Copyright
Result
Verify results
Page
Verify
Verify Distributed Device creation
Not in a Storage View
© Copyright
Create a Distributed Device with the storage-tool compose

Command
storage-tool compose command is one command that will create a Virtual

Volume and export it to a selected host.
Command Details
Command options
Exampl
The storage-tool compose command can be used as a simple way to create a Virtual
Volume. This one command creates the virtual volume on top of the specified storage-volumes,
building all intermediate extents, local, and distributed devices as necessary. Storage-volumes
from each cluster may be claimed but must be unused.
The command allows the user to add the virtual volumes created to both a consistency group
and a storage view. They must exist already. The example displays adding a virtual volume
named new-vv to a storage view named my-view and a consistency group named my-CG.
Page
Create Distributed Devices using VPLEX CLI
Identify Storage Volumes
The ll -p /**/storage-volumes command will list the Storage Volumes by

VPLEX Cluster. Identify Storage Volumes that are claimed.
© Copyright
Create Extents
Here we have created an Extent.
Page
View Extents
The ll -p /**/extents command lists the extents on both VPLEX Clusters.
© Copyright
Create Local Devices
Here a local Device has been created. Notice it is a RAID-1 Device.
Page
Identify Local Devices
This VPLEX CLI command displays all the Devices located at each cluster.
© Copyright
Create Distributed Devices
The CLI command ds dd create is used to create a Distributed Device. The

Devices at each cluster, along with the source, are defined.
Page
View Distributed Devices
Displayed are the properties of a Distributed Device. Notice that the rebuilding
process is still ongoing.
© Copyright
Virtual Volume Create
Once a Distributed Device is created a Virtual Volume can be added. Also

displayed is the show-use-hierarchy command.
The best practice for creating a Distributed Device with the CLI is to
specify the source-leg. when this option is used mirror legs
automatically synchronize.
Page
VPLEX Consistency Group Creation
Here are the steps to create a Consistency Group using Unisphere for VPLEX.
Launch the Wizard
Navigate to Provision Storage>Distributed Storage and view by Consistency Groups.
© Copyright
Steps One and Two
A name for the new Consistency Group can be entered. Step two allows the
selection of a Rule Set.
Enter a name for the new group. Select a Rule Set for the group.
Page
Steps Three and Four
These steps in the wizard will add existing Virtual Volumes to the Consistency
Group. Next, a review of the previous selections is performed.
© Copyright
Step Five
Step five displays the results. These include the group name, detach rule, and the
Virtual Volumes added.
Page
View Consistency Group
Here are the attributes for the Consistency Group.
© Copyright
Distributed Device Failure Scenarios
Distributed Device and Consistency Group Failure Properties
Detach Rules
Two levels of detach rules:
 Detach rules associated with individual Distributed Devices

Every Distributed Device must have a detach rule. There are two device-level detach rules:
 cluster-1-detaches - Cluster-1 is the preferred cluster. During a cluster failure or inter-

cluster link outage, cluster-1 continues I/O and cluster-2 suspends I/O. If Cluster-1 fails
there will be no access.
 cluster-2-detaches Cluster-2 is the preferred cluster. During a cluster failure or inter-

cluster link outage, cluster-2 continues I/O and cluster-1 suspends I/O. f Cluster-2 fails
there will be no access.
 Detach rules associated with Consistency Groups
Every Consistency Group has a detach rule that applies to all members in the Consistency
Group. If a distributed device is a member of a Consistency Group, the detach rule of the
Consistency Group overrides the detach rule configured for the device. Here are the detach
rules:
 winner cluster-name delay seconds - The cluster specified by cluster-name is

declared the preferred cluster if a failure lasts more than the number of seconds
specified by seconds.
 no-automatic-winner - The consistency group does not select the preferred cluster.
The detach rules of the member devices determine the preferred cluster for that device.
Page
Cluster Witness
VPLEX Witness helps VPLEX Metro configurations automate the response to

cluster failures and inter-cluster link outages. VPLEX Witness connects to both
VPLEX clusters over the management IP network.
When VPLEX Witness is not deployed, detach rules determine at which cluster I/O
continues during a cluster failure or inter-cluster link outage. VPLEX Cluster
Witness does not guide Consistency Groups with the no-automatic-winner detach
rule.
© Copyright
Inter-cluster WAN Failure
An inter-cluster WAN communication failure will break the ability for Distributed
Devices to sustain I/O synchronization on local devices across clusters.
LAN
Cluster-1 Partition Failure Cluster-2
(Loss of WAN Links)
 WAN failure will cause all distributed devices to partition.
Page
WAN Connectivity Outage
The connectivity validate-wan-com command shows that there is no WAN

connectivity from cluster-1 to cluster-2. All WAN connections are down.
VPlexcli:/> connectivity validate-wan-com
connectivity: NONE
fc-port-group-3 - FAIL - No connectivity was found from any com port.

© Copyright
System Status
From the dashboard we can see the WAN failure and Storage View error at cluster-
2.
Error due to Cluster-1 is

the Winning Cluster
Page
Wan Restore
Command details - The WAN Connectivity has been restored (after the delay timer expired).
From the VPLEX CLI, we notice that the Distributed Device needs a “resume” at the losing
cluster.
© Copyright
Check Status
Storage Views - Here we see the Storage Views for cluster-1 and cluster-2. Cluster-1 is the
winning cluster in this example. Notice the operational status of the two Storage Views.
Winning Cluster Losing Cluster
Page
Device Status
Use the ll command to view the status and attributes of the Distributed Devices.
© Copyright
Resume Distributed Device
Here we see the command device resume-link-up to resume the Virtual

Volume. This can also be done using the device at the losing cluster. This
command only works once WAN connectivity is restored.
device resume-link-up --virtual-volumes Student_1_VV -force
ll
Page
Volume Expansion and Protection
Virtual Volume Expansion
Concatenation (RAID-
Storage-Volume
C)
VPLEX expands The virtual volume

the underlying is expanded by
storage-volume adding only
(Prior to this step specified extents or
the LUN presented devices, as required
to VPLEX must be
expanded on the
array)
© Copyright
Expansion Types
A VPLEX Virtual Volume can be expanded by two methods. They are Storage Volume
expansion or concatenation. If the volume type supports expansion, VPLEX detects the
capacity gained by expansion. Now you can identify the available expansion method. The
Storage Volume method is always preferred. Possible values for the expansion-method
attribute are:
 Storage Volume
o VPLEX makes use of the underlying storage-volume expansion
 Concatenation
o A Virtual Volume is expanded by adding only specified Extents or

Devices
Page
Procedure for Volume Expansion
When expanding a Virtual Volume, the first step is to determine volume expansion
method. The method available is determined by the underlying Device.
The expansion-method attribute can be listed using the CLI or Unisphere.
Determine volume expansion method
Storage-volume Concatenation (or RAID-C expansion)
Check expansion prerequisites
Migration or rebuilds Non-Disruptive Health-check

RecoverPoint-enabled
in progress? Upgrade in progress? command reports Metadata volumes?
consistency group?
problems?
Expand the virtual volume
VPlexcli Unisphere for VPLEX
Virtual Volume expansion prerequisites:
Volumes cannot be expanded if any of the following conditions are true:
 Migration or rebuilding is occurring
 Upgrade is in progress
 health-check has errors
 Volume to expand belongs to a RecoverPoint-enabled Consistency Group
© Copyright
Storage Volume Expansion Topologies
The storage volume method of expansion supports simple expansion on a variety

of device geometries.
Virtual Virtual Virtual

Volum Volum Volum
RAID-1
Device 1 Distributed RAID-1 device
Device
Extent 1 Extent 1 Extent 2 Extent 1 Extent 2
Storage Storage Storage Storage Storage

Volume 1 Volume Volume 2 Volume 1 Volume 2
LUN 1 LUN 1 LUN 2 LUN 1 LUN 2
Storage Array Storage Array 1 Storage Array 2 Storage Array 1 Storage Array 2
Single Cluster Cluster-1 Cluster-2
1:1 Virtual Volume to Storage
Volume Dual-legged RAID-1 Distributed RAID-1
Storage Volume expansion criteria
The VPLEX Virtual Volume geometry must meet one of the following criteria:
 Mapped 1:1 to the underlying Storage Volume
 A multi-legged RAID-1 or RAID-0 volume and each of its smallest extents is

mapped 1:1 to a back end Storage Volume
 A Distributed Raid-1 Device where the mirror leg at each cluster is mapped 1:1
to the underlying Storage Volume.
There is a maximum initialization processes can run concurrently per cluster. See the Release
Notes for the current limit.
Page
Storage-volume Expansion Unisphere
This is an example of volume expansion through Unisphere for VPLEX.
10 GB
To begin, you can list the expandable-capacity attribute (in the CLI) or the Expandable By
field (in the GUI) to plan capacity of your back-end storage. When using Unisphere, click on the
Virtual Volume name to display the properties of the Virtual Volume you want to expand. For
Virtual Volumes that can be expanded using the Storage-volume method, the Expandable By
attribute is the capacity added to the back-end storage volume, but not yet exposed to the host
by the Virtual Volume. A value of zero indicates that there is no expandable-capacity for the
volume. A non-zero value indicates the capacity available to expand. Here are the steps to
perform Storage Volume expansion:
1. Identify the underling Storage Volume for the Virtual Volume to be expanded.
2. Connect to the array and expand the volume presented to VPLEX.
3. Rediscover the array from VPLEX.
4. Verify the Virtual Volume now has expandable capacity.
© Copyright
Storage-volume Expansion Procedure with CLI
The commands to be used to expand a volume using the storage-volume method:
Starting Capacity
Steps with the CLI for Storage Volume Expansion:
 Use ll <virtual volume> to determine method and capacity

 Use the expand command to perform the expansion.
Page
Concatenation (RAID-C) Expansion Method
Concatenation Expansion
Before Expansion After Expansion
Virtual Virtual
Volume Volume
2
GB
Device 1
RAID-C
1G 1G
Device 1 Device 1 Device 2

Device 2
<date>
Some devices do not support the storage volume method of expansion. In this case, use the
concatenation method. The concatenation method expands the virtual volume by adding only
specified extents or devices.
Top-level Devices can be expanded without disruption to a host. A top-level Device can be
expanded using this method which will add another Device to the first. This creates a new top-
level device of RAID-C type. The device to be appended can be any type of device provided it
is not mapped to a Virtual Volume. It is best practice to use the same geometry as the original
device.
After the expansion is complete, the original mapped Device has been converted into type
RAID-C. This contains the original Device with the date appended to the Device name.
© Copyright
Concatenation Expansion – with CLI
CLI Expansion Method - The expansion begins with a 4 GB virtual volume, and then we
concatenated a 4 GB device to expand the virtual volume to 8 GB by using the virtual-volume
expand command. Confirm the expansion with ll.
Page
Graphic View of Concatenation Expansion
This figure is a graphical representation of the virtual volume shown on the

previous page.
This is the new 8 GB RAID-C

Device
This is the original device. It was

This is the new 4 GB device. built from a 4 GB Extent (which came
That was added from a 4 GB slice of an 8 GB Storage
Device)
This figure is a graphical representation of the virtual volume shown on the

previous page.
© Copyright
Volume Protection – Add a Local Mirror
Choose Virtual Volume
The first step when adding local mirror protection is to select a Virtual Volume and
launch the wizard.
Select Virtual Volume
Page
Select Devices
Based on the Virtual Volumes selected, the devices to mirror are automatically
selected.
© Copyright
Select Mirrors
Step two will select the target Device. It must be on the same cluster as the source
and the same size or larger.
Page
Complete
The job results are shown after a review, which is not shown.
© Copyright
Volume Protection – Attach a Local Mirror with the CLI
The device attach mirror command is used to add a local device as a mirror
leg.
The show-use-hierarchy command displays the complete usage hierarchy for

a storage element. This view is from the top-level element down to the storage-
array.
Note: This same command can be used to create a Distributed Device.

The source and target Devices are located on different VPLEX Clusters.
Page
Adding a Remote Mirror
Virtual Volume Selection
The steps for adding remote RAID-1 protection to a Virtual Volume are similar to
adding a local mirror.
The local Device may be any type geometry, but the size of the source must be
equal to or less than the target.
Select a Virtual Volume
© Copyright
Add Remote Mirror Wizard
The Device to mirror will be selected automatically. Select a Device located in the remote cluster. The
The selection is based on the previously selected best practice it to select a target that is the same size
Virtual Volume. as the source.
Page
Steps Three and Four
Because the new Device is now a Distributed Device Review the selections
a Consistency Group or Rule-set must be selected.
© Copyright
Completion
Here we see the adding of a remote mirror was successful.
Page
Device Map
Here is an example of the Device Map. Notice the new Top-level Device. There is
also a new Virtual Volume in Cluster-1, which contained the added Device.
© Copyright
Data Protection with RecoverPoint
VPLEX with RecoverPoint Protection
Sample VPLEX and RecoverPoint Configuration
RPA Cluster RPA Cluster
VPLEX Unity
This example is a VPLEX Local configuration with RecoverPoint providing both

local and remote Virtual Volume Protection. RecoverPoint provides a local copy,
which is a Virtual Volume on the VPLEX. This configuration also provides a remote
copy. In this example, the data it is stored on a Dell EMC Unity array.
RecoverPoint Details - RecoverPoint provides a software embedded “Write-Splitter”. This

software is built into the VPLEX code and is enabled when RecoverPoint is implemented. The
splitter is also part of other Dell EMC supported arrays. The splitter copies incoming write I/Os
from the host for selected Virtual Volumes. These changes are sent to the RecoverPoint
Appliance (RPA). The RPA updates a Journal which stores all the changes made. Each copy
requires a Journal. In this example, Journals are Virtual Volumes on the VPLEX. These Virtual
Volumes must be in a Storage View along with the RecoverPoint Appliances. They are also
required to be in a RecoverPoint enabled Consistency Group. In this example there is a local
copy, which is a Virtual Volume on the VPLEX. This Virtual Volume, and all others used as
copies, are required to be in a Storage View and a Consistency Group. This includes the Virtual
Volumes that are Production volumes seen by hosts. The Storage View must only include the
RPA initiator ports. No other host should be part of this view. The Consistency Group must be
RecoverPoint enabled. It is required to have separate groups for Journals and copies.
Page
Procedure for Adding RecoverPoint Protection
Here is an overview of the steps that are required to add RecoverPoint protection
to VPLEX Virtual Volumes. Each step is listed as to where, RecoverPoint or
VPLEX, it is performed.
VPLEX VPLEX VPLEX VPLEX
Import RecoverPoint Add RecoverPoint Register Create RecoverPoint

certificate in VPLEX cluster to VPLEX RecoverPoint storage view
VPLEX RecoverPoint RecoverPoint
Validate Create Add the VPLEX

Create VPLEX
RecoverPoint in RecoverPoint Splitter to
consistency groups
VPLEX Consistency Groups RecoverPoint
VPLEX
© Copyright
Importing RecoverPoint Certificate
Certificates are required for VPLEX and RecoverPoint to obtain management

information about volume types. The RecoverPoint certificate is imported into
VPLEX, and the VPLEX certificate is imported in RecoverPoint.
Page
Create VPLEX Storage View for RecoverPoint
Register Initiators
A RecoverPoint Cluster contains two to eight RecoverPoint Appliances (RPA).

Each RPA has 4 initiator ports.
Set the Host Type to recoverpoint for each initiator.
© Copyright
Create Storage View
A Storage View for the RecoverPoint Cluster must contain the following:
 All RPA ports

– A RecoverPoint Cluster has 2-8 Appliances
– Each appliance has 4 ports
 VPLEX Front-End ports used by the RecoverPoint cluster
 All Virtual Volumes required by the RecoverPoint Cluster
Page
Adding RecoverPoint Cluster to VPLEX
Use the rp rpa-cluster add command on the VPLEX and add the clusters
that are local to VPLEX Metro to the VPLEX system. Ensure that you specify which
VPLEX cluster is the local cluster.
Add RecoverPoint Cluster to each VPLEX Cluster (Metro)
© Copyright
Create VPLEX Consistency Group
Create a RecoverPoint enabled Consistency Group for all the Virtual Volumes
required by the RecoverPoint Cluster.
Use separate Consistency Groups for Local and Distributed Virtual Volumes.
Consistency Groups for Consistency group for

replication volumes journals and repository
Separate local and Distributed All Journals must be in the same

Virtual Volumes Consistency Group
Separate local copy volumes from Add the RecoverPoint Repository

production volume if they reside on the Volume ( if required) to the same group
same clusters as the Journals
Page
VPLEX Data Mobility
VPLEX Data Mobility
Data Mobility Use Cases
Without the storage virtualization that VPLEX offers, migrating data from one array
to another is difficult. This procedure would require professional services to plan
and implement the migration.
VPLEX can handle many data mobility needs and migrate data from one array to
another with minimal effort and no planned downtime.
IT needs to move data. VPLEX gets your data wherever you want with
no planned downtime
Tech refresh Datacenter moves

Consolidation
Load balancing
© Copyright
VPLEX Data Mobility
Data Mobility Overview
VPLEX Data Mobility moves data from one Extent or Device to another. The Virtual
Volume, which can be in a Storage View, remains unchanged. The "volume
identifier" remains unchanged. This allows moving data without host or application
disruption.
Virtual
Volume
VPLEX Back End Device

Mobility operations are
transparent to host
Data Mobility Types:
 Extent
– Within a Cluster
 Device
– Within a Cluster or between Clusters
Page
VPLEX Data Mobility
Data Mobility Overview
WARNING: Device migrations are not recommended between clusters. All

device migrations are synchronous. If there is I/O to the devices being
migrated, and latency to the target cluster is equal to or greater than 5 ms,
then significant performance degradation could occur.
VPLEX does not remove the Data from the old Storage Volume.
© Copyright
VPLEX Data Mobility
Extent Mobility
Virtual
Volume
Device
Sourc Target
Extent Mov Extent
Storage Storage
Volume Volume
Storage Storage
Array Array
Extent mobility is a VPLEX mechanism to move all data from a source Extent to a
target Extent.
 Moves data from a source Extent to a target Extent

 Non-disruptive to the host
 Within a cluster only
 When committed, frees the original Extent for reuse
 Foundation of non-disruptive data mobility across or within storage arrays
Page
VPLEX Data Mobility
Device Mobility
Virtual
Volume
Temporary RAID-
1 Device
Mirror Mirror
Leg Leg
Device Move Device

Source Target
Extent
Extent Extent
Storage
Storage Storage
Volume
Volume Volume
© Copyright
VPLEX Data Mobility
Device Mobility Details:
 Moves data from a source device to a target device

 Non-disruptive for host
 When committed, frees the source device and underlying extent(s)
 Local or remote
 Local - within a cluster
 Remote - across clusters
 Best method for array migrations
Page
VPLEX Data Mobility
General Procedure to Perform Data Migration
Optional
Create and check a Start the Monitor the

Pause, resume, or
migration plan migration migration
cancel the migration
progress
Batch Migration Only
Remove the Commit the

Clean up
record of the migration
migration
Optional
Both GUI and CLI commands have options to:

start/pause/resume/cancel/commit/clean/remove
© Copyright
VPLEX Data Mobility
VPLEX Data Migration Considerations
Device migrations between distributed devices are not supported
Devices must be removed from consistency groups before being migrated
Target device must be the same size or larger than the source device or extent
Target devices must not have existing virtual volumes
Migration supported for “Thin Enabled” virtual volumes
Page
VPLEX Data Mobility
VPLEX Batch Migrations
Batch migrations migrate multiple Extents or Devices. Create batch migrations to

automate repetitive tasks.
Batched extent migrations Batched device migrations
Migrate arrays within Migrate to dissimilar

the same cluster arrays
Same number of LUNs in the Migrate devices between

source and destination clusters in a VPLEX Metro
Identical capacities in the

source and destination
© Copyright
VPLEX Data Mobility
Extent Mobility
Starting the Wizard
Here is a map of the components of a Virtual Volume prior to migration. Note the names of the Storage Volumes and Extents.
Page
VPLEX Data Mobility
Using the Mobility menu, select Move Data Within Cluster for Extent Mobility.
© Copyright
VPLEX Data Mobility
Extent Mobility Wizard
Select the desired cluster Select a Storage Volume, this step is optional and acts
as a filter. The Extent on this Storage Volume will be
selected in the next step.
Page
VPLEX Data Mobility
Extent Selection
Here we select an Extent to migrate.
Select Target
Based on the previous selections, a target Extent is presented for selection. The
Auto-Generate Mappings can be selected to automatically choose the target.
© Copyright
VPLEX Data Mobility
Review
In step 5, configure the job name and set the rate to perform the migration.
Page
VPLEX Data Mobility
Monitor Extent Mobility Jobs
Once the transfer is complete, the job can be Here is a map example for the completed migration.
committed. At this time it also can be canceled. Compare it to the previous map.
Committed jobs cannot be undone.
© Copyright
VPLEX Data Mobility
Page
VPLEX Data Mobility
Device Mobility Wizard
Launch the Wizard
Devices can be migrated either locally or to a remote cluster.
Select Create
© Copyright
VPLEX Data Mobility
Select Cluster
Select Cluster Select Virtual Volume
Page
VPLEX Data Mobility
Select Source and Target
© Copyright
VPLEX Data Mobility
Review
Page
VPLEX Data Mobility
Commit
Once the job transfer is complete it can be committed.
© Copyright
VPLEX Data Mobility
Complete Device Mobility
Once the job transfer is complete it can be committed. After the migration is
complete, the commit step detaches the source leg of the RAID 1 and removes the
RAID 1. The Virtual Volume, Device, or Extent is identical to the one before the
migration except that the source Device/extent is replaced with the target
Device/Extent.
Page
VPLEX Data Mobility
Mobility Operations Through CLI
Four Data Mobility Operations
VPlexcli:/data-migrations/device-migrations> dm migration start -- name migrate_012 --from device_012 --to device_012a -transfersize 12M
VPlexcli:/> ls data-migrations/device-migrations/ migrate_012

...
Use the ls command to
start-time Fri April 8 13:32:23 MDT 2016
display the migration's
status in progress status.
...
VPlexcli:/data-migrations/device-migrations> dm migration commit -- force --migrations migrate_012

Committed 1 data migration(s) out of 1 requested migration(s).
VPlexcli:/data-migrations/device-migrations> dm migration clean -- force --migrations migrate_012

Cleaned 1 data migration(s) out of 1 requested migration(s).
VPlexcli:/data-migrations/device-migrations> dm migration remove -- force --migrations migrate_012
Removed 1 data migration(s) out of 1 requested migration(s).
RAID-1
Source Device or Extent Destination Device or Extent
Mov
There are four basic operations involved in moving Extents or Devices:
The start operation first creates a RAID 1 device on top of the source Device. It specifies the
source device as one of its legs and the destination Device as the other leg. It then copies the
source Device’s data to the destination’s Device or Extent. This operation can be canceled as
long as it is not committed.
The commit operation removes the pointer to the source leg. At this point in time the
destination Device is the only Device accessible through the Virtual Volume.
The clean operation breaks the source Device down all the way to the Storage Volume level.
The Storage Volume is unclaimed after this operation if there are no other Extents configured
for this Storage Volume. Data mobility operations can also be paused and resumed before the
commit operation. It may be beneficial to pause mobility operations during daytime hours.
The remove operation will remove the record of canceled or committed data migrations.
© Copyright
VPLEX Data Mobility
Batched Mobility
Batched Mobility Details
 Allows the batching of extent or device mobility

jobs
 Enables convenient large scale mobility operations
 Makes use of a Plan File
 Up to 25 jobs at a time
Batch migrations are run as batch jobs from reusable batch migration plan files. Migration plan
files are created using the create-plan command. A single batch migration plan can be either
for Devices or Extents, but not both. Batched mobility provides the ability to script large-scale
mobility operations without having to specify individual extent-by-extent or device-by-device
mobility jobs. Batched mobility can only be performed in the CLI. Batch migrations must follow
the same rules as individual migrations.
Use batch migrations to:
 Retire storage arrays and bring new ones online
 Migrate devices to a different class of storage array
There are two additional steps to prepare for a batch migration:
 Create a batch migration plan file using batch-migrate create-plan
 Test the batch migration plan file using batch-migrate check-plan.
Page
VPLEX Data Mobility
Create and Check the Batched Mobility Plan
The batch-migrate create-plan command creates a migration plan using

the specified sources and targets.
© Copyright
VPLEX Data Mobility
Start and Cancel Batched Jobs
Use the batch-migrate start --transfer-size [40 KB-128 MB]

<filename> command to start the specified batch migration job.
Use the batch-migrate cancel --file <filename> command to cancel.
VPlexcli:/clusters/cluster-1/devices> batch-migrate start --file=migrate.txt --transfer-size 2M
Started device migration 'BR0_0'.
Started 3 of 3 migrations.
Job Start
Job Cancel
VPlexcli:/data-migrations/device-migrations> batch-migrate cancel --file migrate.txt
• Transfer-size can be as small 40 K, as large as 128 M, and must be a multiple of

4 K. The default recommended value is 128 K.
• A larger transfer-size results in higher performance for the migration, but may
negatively impact front-end I/O.
Page
VPLEX Data Mobility
Pause and Resume Batched Jobs
Pause an active batch migration to release bandwidth for host I/O during periods of
peak traffic. Resume the batch migration during periods of low I/O.
VPlexcli:/clusters/cluster-1/devices> batch-migrate pause --file=migrate.txt
WARNING: Failed to pause migration BR0_1 : Evaluation of <<dm migration pause -m /data-
migrations/device-migrations/BR0_1>> failed.
Unable to pause the given data migration(s).
Could not pause migration 'BR0_1' because it is not in-progress.
Paused 2 of 3 migrations Job Pause
Job Resume
VPlexcli:/clusters/cluster-1/devices> batch-migrate resume --file=migrate.txt
WARNING: Failed to resume migration BR0_1 : Evaluation of <<dm migration resume - m /data-
migrations/device-migrations/BR0_1>> failed.
Unable to resume the given data migration(s).
Could not resume migration 'BR0_1' because it is not paused.
Resumed 2 of 3 migrations
© Copyright
VPLEX Data Mobility
Monitor Batched Mobility Jobs
Use the batch-migrate summary <filename> --verbose command to

monitor the progress of the specified batch migration.If more than 25 migrations are
active at the same time, they are queued, their status is displayed as in-progress.
Page
VPLEX Data Mobility
Commit, Clean and Remove Batch Jobs
After the migration is complete, the commit step detaches the source leg of the
RAID 1 and then removes it.
VPlexcli:/> batch-migrate commit –-file migrate.txt
VPlexcli:/> batch-migrate clean --rename-targets --file migrate.txt

Using migration plan file /temp/migration_plans/migrate.txt for cleanup phase.
0: Deleted source extent
/clusters/cluster-1/devices/R20061115_Symm2264_010, unclaimed its disks Symm2264_010
1: Deleted source extent
/clusters/cluster-1/extents/R20061115_Symm2264_011, unclaimed its disks Symm2264_011
VPlexcli:/> batch-migrate remove /data-migrations/device-migrations --file migrate.txt
After the migration is complete, the commit step detaches the source leg of the RAID 1 and
then removes it. The Virtual Volume, Device, or Extent is identical to the one before the
migration except that the source Device/Extent is replaced with the target Device/Extent. A
migration must be committed in order to be cleaned.
When the batch migration is 100% complete, use batch-migrate commit <filename>. Next,
run the clean command to dismantle the source device down to its Storage Volume.
Remove the migration record only if the migration has been committed or canceled. Migration
records are in the /data-migrations/device-migrations context.
© Copyright
Role-Based-Access-Control
Supported RBAC Roles
Role-name Description Shell Access
service  Used to configure Always
VPLEX (Dell
Technologies Service
personnel only)
securityadmin  User management Always
(only admins should

use this)
 Can also perform
vplexuser tasks
vplexuser  Standard VPLEX Controlled by admin
management -
provisioning
 Cannot do any
configuration (service
tasks) or account
management
readonly  Monitoring only - no Never
changes allowed
 CLI or REST
monitoring scripts can
use this account
Page
Accounts with “vplexuser” Role
The vplexuser role is for accounts created by the admin or accessed via LDAP.
Hos
vplexus
Automated Tasks
vplexuser
Storage
Arrays
Standard
VPLEX User
Standard VPLEX management tasks:
 Provision
 Monitoring
 Mobility
Cannot create accounts or perform service tasks.
© Copyright
Accounts with “readonly” Role
Hos
readonly
Automated Tasks
readonly
Storage
Arrays
VPLEX
Monitoring
User
The readonly role allows automated monitoring tools read only access to VPLEX.
Accounts with the readonly role can be created.
Page
RBAC - View Account Role Information
Navigate to the /management-server/users/local context to view the user-name

and role-name for any user. You can also see whether or not a particular user is
allowed shell access.
© Copyright
RBAC - Change Role for User
The VPLEX administrator can change the role and shell access for any account
with a vplexuser or readonly role.
 Currently only accounts with readonly and vplexuser role are allowed to be
changed.
 Role for service and admin account cannot be changed
Page
Shell Access Control - Enabled /Disabled
Shell Access Details
Service and admin accounts Accounts with vplexuser and readonly Roles
login as: admin login as: newuser1
Using keyboard-interactive authentication. Using keyboard-interactive authentication.
Password:
Password:
service@ManagementServer:~> vplexcli Trying ::1...
Trying ::1... Connected to localhost.

Access
Escape character is '^]'.
Connected to localhost. to Linux
shell creating logfile:/var/log/VPlex/cli/session.log
Escape character is '^]'.
VPlexcli:/>
creating logfile:/var/log/VPlex/cli/session.log
User with shell access = false will
directly log into VPLEX CLI
VPlexcli:/>
These two example screenshots show the difference when logging in with a role that has shell
access, and logging in with a role that does not have shell access. The example on the left has
shell access. Notice in the example on the right that newuser1 does not have shell access. So
after entering the password, newuser1 is put directly into Vplexcli.
Shell access can be enabled or disabled for any user assigned the vplexuser role.
When restricted users exit the VPLEX CLI by any method, they will also exit from the shell.
There is no way for a restricted user to exit VPLEX CLI and get to the shell prompt.
© Copyright
If Restricted Shell Access - SCP Using "share" Folder
There is a restricted “share” folder accessible through the VPLEX CLI.
 Use the ll command in the /management-server/share/out/collect-

diagnostics context to see the files that are stored.
 To copy a diagnostics file from the share folder, we use the command: scp
newuser@<ipaddress>:/collect-diagnostics/FNMXXXXXXX-c1-
diag-xxxx.tar.gz.
Page
VPLEX Support Integration
Configuring SRS Gateway
Secure Remote Services (SRS) is a two-way remote connection between Dell EMC
Customer Service and supported products and solutions.
SRS Virtual Edition (VE) runs on a customer-supplied Enterprise VMware or Hyper-

V instance.
Required Information from VPLEX
• IP address of the management server
• VPLEX serial number (Top Level Assembly #)
• Site ID
SSL tunnel - TLS with RSA key exchange AES-

256 with SHA1 encryption
SRS Gateway
Administrator ← Proactive Remote Monitoring →

Dell Support
© Copyright
SNMP Overview
 SNMP can retrieve performance statistics for a single cluster

 One-time SNMP agent configuration is necessary
 Call-home event notifications sent
 Third-party management station often used

SNMP Management
Station
SNMP Polling Polling Information
VPLEX Management Server or

SNMP Traps
VPLEX Engines
SN Cluster-1
SNMP version 2c supported
Page
Supported SNMP Polling Commands
The relevant commands for SNMP polling are displayed here:
 VPLEX CLI
 snmp-agent configure
 Configures the SNMP Agent
 One time configuration
 Executed in the VPLEX CLI
 SNMP Management Station
 SNMPGET
 gets the most recent statistics for the OID specified from each director
 SNMPGETNEXT
 gets the most recent statistics for the next OID from each director
 SNMPGETBULK
 gets all SNMP statistics from each director
© Copyright
Supported SNMP Trap Commands
SNMP Traps are configured in the VPLEX CLI.
 notifications snmp-trap create
 Creates an SNMP trap sink for call-home events

 set remote host
 Configures the trap destination

 set started true
 Begins SNMP trap monitor

 notifications snmp-trap destroy
 Deletes one or multiple SNMP traps
Page
Using LDAPS for User Accounts
Configure user accounts one of two ways:
 Use an external LDAPS/AD (Lightweight Directory Access Protocol/Active

Directory)
 Configure locally on VPLEX
IP
Admin VPLEX Management Server or

MMCS-A
CERT
LDAPs Server
© Copyright
LDAPS Configuration on VPLEX
VPLEX provides the option to configure Lightweight Directory Access Protocol

Secure (or LDAPS) certificates which is the more commonly used form of LDAP
and is more secure. The certificates are files that come from the LDAPS server and
are stored on the VPLEX, which creates a trust mechanism between the two
systems.
VPlexcli:/> authentication directory-service configure -d 1 -t 2 -I 10.31.50.59 /
--server-name=“linux-72.s3.site" -b “dc=emc,dc=com" /
-r “ou=vplex,dc=emc,dc=com" -n “cn=Administrator,dc=emc,dc=com" /
-l “/opt/emc/Bplex/cert.pem" -p
...
VPlexcli:/>
-d for directory type, 1= OpenLDAP, 2= Active Directory
-t for connection type, 1= LDAP. 2=LDAPS
-i for IP address of the certificate server
--server-name for the name of the LDAPS server
-b for the base distinguished name of the directory server
-r for the user search path, the distinguished name of the node at which to begin
user searches in the directory server
-n for the bind distinguished name of the directory server
-l the path to the certificate file
-p for Password of Bind Distinguished Name. You are prompted for the
password
Page
VPLEX Monitoring Concepts
Monitor Clusters
The cluster summary command is a good command to run to monitor VPLEX

clusters.
VPlexcli:/> cluster summary
Clusters:
Name Cluster ID TLA Connected Expelled Operational Status Health State
--------- ---------- -------------- --------- -------- ------------------ ------------
cluster-1 1 FNM00161800086 true false ok ok
cluster-2 2 FNM00161800085 true false ok ok
Islands:
Island ID Clusters
--------- --------------------
1 cluster-1, cluster-2
Clusters in a VPLEX Metro will share the same Island ID.
VPlexcli:/> cluster status
operational-status: ok
transitioning-indications:
transitioning-progress:
health-state: ok
health-indications:
local-com: ok
...
© Copyright
VPLEX Monitoring
Validate VPN Status

The vpn status command verifies the VPN connection between management
servers.
Page
Export Port Summary
This command allows the user to check the status of VPLEX Front-End ports. xxxx
© Copyright
Director Connectivity
The director uptime command displays the amount of time a director has
been online. The connectivity director <director name> command will
display all ports and storage array LUNs masked to the specified director.
Page
Monitor Users
The sessions command displays information on users who are logged into the
VPLEX Management Console.
© Copyright
Monitor System Volumes
A metadata volume can be monitored by listing the system-volumes directory. If the

metadata volume is having trouble, it will be shown here. Note that a cluster will not
allow configuration changes to the metadata volume if both mirror legs of a
configured metadata volume have failed. I/O to existing Virtual Volumes will not be
affected.
Page
VPLEX Monitoring
Default Report – capacity-arrays
This monitor generates a capacity report for all the storage in a VPLEX system, grouped by
storage arrays. It requires:
 All Storage Volumes in a storage array have the same tier value
 The tier is indicated in the Storage Volume name
Tier IDs are required to determine the tier of a Storage Volume/storage array. Storage Volumes
that do not contain any of the specified IDs are given the tier value 'no-tier'. The report is
separated into two parts:
 Local storage - Storage Volumes where the data is physically located at one
site only.
 Shared storage - Distributed and remote Virtual Volumes
© Copyright
Default Report – capacity-clusters
Information generated from the report capacity-clusters monitor.
Page
Default Report – capacity-hosts
Information generated by this monitor includes the number of views, total exported
capacity in GB, and the number of exported virtual volumes per cluster.
© Copyright
Monitor VPLEX Object Constructs
The show-use-hierarchy command drills from the specified target up to the

top-level volume and down to the storage-array.
Details - The command will detect sliced elements, drill up through all slices, and indicate in
the output that slices were detected. The original target is highlighted in the output. You can
specify meta, logging, and virtual volumes, local and distributed devices, extents, storage-
volumes, or logical-units on a single command line.
Page
Monitor Storage Volume Details
This command provides additional details on storage volumes.
© Copyright
VPLEX Monitoring
Monitor Storage Volumes
The storage-volume-summary command provides information on the number

of used, claimed, and unhealthy storage volumes. It also displays the number of
meta-data volumes, the storage volume vendor, and the total capacity of storage
volumes.
Page
Extent Details
The ll/**/extents command provides detailed information on the extents in a

particular cluster. It will display , among other details, the use details. this provides
information on an Extent
© Copyright
Local Device Details
This command provides additional details on Devices, including the operational

status and health state, block count, block size, geometry, and the Virtual Volume
using the Device.
Page
Virtual Volume Details
This command provides additional details on Virtual Volumes. These details

include Virtual Volume operational status and health state, block count and block
size, Virtual Volume capacity, locality, and Virtual Volume supporting Devices.
© Copyright
Virtual Volume Summary
The virtual-volume summary command displays information about the Virtual

Volumes in a cluster. This information includes the Virtual Volume health state,
locality, cache mode, capacity, and total capacity.
Page
VPLEX Monitoring
Storage View Details
The ll storage-views command displays detailed information on Storage

Views. This includes the operational status of a view, initiator ports, Virtual
Volumes, and VPLEX Front-end ports that are part of the view. It also displays the
external LUN ID of a Virtual Volume.
Storage View find
The export storage-view find-unmapped-volumes <cluster>

command will display the virtual volumes that are un-exported (not placed in a
storage view).
The export storage-view map <view> will display virtual volumes that are
part of the Storage View and their corresponding local device IDs.
© Copyright
Monitor Distributed Devices
These commands provide information about Distributed Devices within a local

cluster or a VPLEX Metro configuration.
Page
VPLEX Performance Monitoring
Overview of VPLEX Performance Monitoring
VPLEX Performance Monitor tool allows for data collection of various statistics from
the VPLEX cluster.
 Unisphere Performance Monitoring Dashboard

 VPLEX CLI commands collect performance statistics
 SNMP (SNMP agent runs on local cluster management server)
VPLEX Performance
Graph
Cluster-1 Cluster-2
Tools for Performance Monitoring
© Copyright 2020 Dell Inc. Page 249

VPLEX CLI Performance Monitors
There are three categories of performance monitors that can be viewed through
VPLEX CLI. All of the VPLEX CLI performance monitors will gather data during a
polling cycle and save it to a File Sink.
Monitor Data
VPLEX Cluster 1
/var/log/VPlex/cli
Three Types for Performance Monitors:
 Perpetual Monitors
– Basic Data - Always Running
 Pre-configured Monitors
– Three per Director
– Must be created
 Custom Monitors
– Created and polling interval set

– Time running is short
Page 250 © Copyright 2020 Dell Inc.

Perpetual Performance Monitor Files
Here we see the contents of the folder where the perpetual monitors are written.
Notice the naming convention of the monitor files. Perpetual monitor files are
collected as part of collect-diagnostics.
Perpetual Performance Monitors details
The Perpetual Performance Monitors are always on, started from system setup, cannot be
modified, disabled, or deleted. The currently open monitor file is capped at 10 MB per director
and up to 10 files are stored (.log, .log.1, .log.2, and so on).
 Always on - cannot be modified, disabled, or deleted
 Records a default set of performance categories every 30 seconds
 One file per director, local to a cluster
 Files written to /var/log/VPlex/cli/
o Naming: …PERPETUAL_vplex_sys_perf_mon.log
 Files capped at 10 MB and rotated up to 10 times
 Files can be copied for offline analysis


Create Pre-configured Monitors
Shown here is an example of the report create-monitors command. Notice

that each pre-configured monitor has one file sink. By default, output files are
located in /var/log/VPlex/cli/reports/ on the management server.
Pre-Configured Monitors


Verify Running Monitors
Use the ll command, as shown here, to verify the running monitors on each
director. Notice that the pre-configured monitors have a period of 0s (zero
seconds). This means automatic polling is disabled. Use the report poll-monitors
command to force the monitors to poll for data and send the data to the associated
file sink.

Manually Poll for Pre-configured Monitors
The monitors created by the report create-monitors command have their

period attribute set to 0 seconds, automatic polling is disabled. use the report
poll-monitors command to force an immediate poll and collection of
performance data.
Output is written to files located in /var/log/VPlex/cli/reports/ on the management

server.

Add Manual Polling to the Scheduled Tasks
To automate polling for our pre-configured monitors, we can schedule a job to run
at specified time(s). The example shown will run the report poll-monitors command
at 1 AM every day.
VPlexcli:/> schedule add –t "0 1 * * *" –c ""report poll-

monitors""
1. Minute
VPlexcli:/> 3. Day of Month 5. Day of Week
2. Hour 4. Month

Custom Monitor Configuration Steps
Here are the general steps to create a Custom Monitor:
1. Determine the type of statistic to collect monitor stat-list

2. Determine how often the monitor should collect statistics.
3. Create a monitor monitor create
4. Add one or more sinks to the monitor monitor add-file-sink
5. Update/Collect statistics
6. Monitor output location /var/log/VPlex/cli/<file name>

Determine the Type of Statistics to Collect
Before creating a monitor, first use the monitor stat-list command to display the
available statistics. There are high-level categories each with subcategories.
Monitoring has no impact on host performance.
Many statistics require a target port or volume to be specified. Output of the monitor
stat-list command identifies which statistics need a target defined.
No Target Required
Target Required

Types of Statistics
Statistics Details
 counters - monotonically increasing value (analogous to a car odometer).

Counters are used to count bytes, operations, and errors. Often reported as a
rate such as counts/second or KB/second.
 readings - instantaneous value (analogous to a car speedometer). Readings
are used to display CPU utilization and memory utilization. Value can change
every sample.
 Period-average - average of a series calculated over the last sample period.
 buckets - histogram of 'bucketized' counts. Buckets are used to track latencies,
determine median, mode, percentiles, minimums and maximums.

Example: How to Create a Custom Monitor
Custom Monitor Details

VPLEX Administration Student Guide

Uploaded by

Copyright:

Available Formats

VPLEX Administration Student Guide

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

VPLEX Administration Student Guide

Uploaded by

Copyright:

Available Formats

VPLEX ADMINISTRATION

VPLEX Ecosystem ............................................................................................................... 8

VPLEX Architecture ................................................................................................. 23

Internal Use - Confidential

VPLEX Management Options .................................................................................. 56

VPLEX Storage Provisioning Concepts ................................................................. 71

Storage Provisioning Methods ............................................................................... 78

Internal Use - Confidential

Storage Volume Encapsulation ............................................................................ 116

VPLEX Distributed Device Concepts.................................................................... 119

VPLEX Distributed Device Configuration ............................................................ 130

Internal Use - Confidential

Distributed Device Failure Scenarios ................................................................... 154

Volume Expansion and Protection ....................................................................... 163

Data Protection with RecoverPoint ...................................................................... 182

VPLEX Data Mobility .............................................................................................. 189

Internal Use - Confidential

Role-Based-Access-Control .................................................................................. 218

VPLEX Support Integration ................................................................................... 225

Internal Use - Confidential

VPLEX Monitoring Concepts ................................................................................ 231

VPLEX Performance Monitoring ........................................................................... 249

Internal Use - Confidential

VPLEX is an important element in the Dell Technologies ecosystem. It is the core

o VPLEX is an important element in the EMC ecosystem and is the core of

o RecoverPoint and RecoverPoint for VMs can provide continuous data

Application Clusters stretched across data centers

70+ Non-Dell EMC Storage platforms

In a worst case scenario, what is the best possible

IT continues as if nothing happened.

Data Mobility Challenges

Traditional methods result in…

• Poor IT resource utilization

• Migration service costs

• Months for Tech Refresh

• Risk of things going wrong

Over time, change to any infrastructure is unavoidable, especially in a cloud

Active-Active Data Center with VPLEX

ACTIVE Continuous data ACTIVE

Stretched host clusters

VPLEX Cluster 1 VPLEX Cluster 2

Simultaneous R/W Non-disruptive data

VPLEX storage virtualization and active-active technology leveraging VPLEX's

Availability with VPLEX

The key component used to enable an active-active data center is VPLEX

 Increased performance/locality of data access - data need not be read from

Virtual Volume Virtual Volume

Distributed Virtual Volume

VPLEX Product Family

VPLEX Local Cluster 1 Cluster 2

VPLEX for ALL Flash

Where and When to Use VPLEX

Continuous Availability Data Mobility

Protect data in the Move data

Use VPLEX for:

 Protect data in the event of disasters or failure of components in your data

VPLEX Use Case - Online Tech Refresh

1. New array is Connected to the SAN

VPLEX Use Case - Storage Pooling