Huawei CloudEngine 6855 Data Center Switch Datasheet
Huawei CloudEngine 6855 Data Center Switch Datasheet
Huawei CloudEngine 6855 Data Center Switch Datasheet
Datasheet
Huawei CloudEngine 6855 series switches are next-generation high-performance and high-density 10GE access
switches designed for data centers (DCs), and provide 40GE/100GE uplink ports.
Overview
● Huawei CloudEngine 6855 series switches are next-generation high-performance and high-density 10GE access switches
designed for DCs.
● CloudEngine 6855 series switches have an advanced hardware structure design, and provide high-density 10GE port access
and 40GE/100GE uplink ports. The switches support extensive DC features and flexible selection of the airflow direction.
● CloudEngine 6855 series switches can work with CloudEngine 16800, 16800-X, or 8800 DC core switches to build an elastic,
virtualized, high-quality, and fully-connected 40GE/100GE data center network (DCN), meeting network requirements of
DCs in the cloud era.
● CloudEngine 6855 series switches provide high-density 10GE access in DCs to build DCN platforms for enterprises and
carriers in the cloud era. The switches can also work as core or aggregation switches on campus networks.
CloudEngine 6855-48XS8CQ
● A CloudEngine 6855 series switch supports up to 48 x 10GE ports, ensuring high-density 10GE server access and smooth
evolution.
● A CloudEngine 6855 series switch supports up to 8 x40GE/ 100GE high-performance QSFP28 ports. Each 100GE QSFP28 port
can be used as one 40GE QSFP+ port or split into four 25GE SFP28 ports or four 10GE SFP+ ports, providing flexibility in
networking. The CloudEngine 6855 switch can connect to the CloudEngine 16800-X, 16800, 9800 or 8800 series switches
through 40GE/100GEuplinks to build a non-blocking network platform.
● CloudEngine 6855 series switches support Multichassis Link Aggregation Group (M-LAG) to implement link aggregation
among multiple devices, improving link reliability from the card level to the device level.
● Switches in an M-LAG all work in active state to share traffic and back up each other, enhancing system reliability.
● Switches in an M-LAG system can be upgraded independently. During the upgrade, other switches in the system take over
traffic forwarding to ensure uninterrupted services.
● M-LAG supports dual-homing to Ethernet, VXLAN, and IP networks, allowing for flexible networking.
⚫ CloudEngine 6855 series switches can work with the industry's mainstream virtualization platforms. When functioning as
high-performance hardware gateways on an overlay network (VXLAN), CloudEngine 6855 series switches can support the
operations of a DC with up to 16 million tenants.
⚫ When functioning as hardware gateways on an overlay network, CloudEngine 6855 series switches can connect to cloud
platforms through open APIs, facilitating unified management of virtual and physical networks.
⚫ The hardware virtualized gateway solution achieves rapid service deployment without having to change the customer
network, protecting customer investments.
⚫ CloudEngine 6855 series switches support Border Gateway Protocol - Ethernet VPN (BGP-EVPN), which can run as the VXLAN
control plane to simplify VXLAN configuration within and between DCs.
⚫ CloudEngine 6855 series switches support NETCONF and can interconnect with iMaster NCE-Fabric.
⚫ CloudEngine 6855 series switches support Ansible — an automatic management and O&M tool — to implement unified
provisioning of physical and virtual networks.
⚫ CloudEngine 6855 series switches are integrated with mainstream cloud platforms (including commercial and open-source
cloud platforms) and third-party controllers, enabling flexible service customization and automatic management.
● CloudEngine 6855 series switches support Zero Touch Provisioning (ZTP). ZTP enables the switches to automatically obtain
and load version files from a file server, freeing network engineers from onsite configuration and deployment. ZTP reduces
labor costs and improves device deployment efficiency.
● ZTP supports embedded script languages and provides them for users through open APIs. DC users can use a familiar
programming language (such as Python) to centrally configure network devices.
● CloudEngine 6855 series switches support telemetry technology to collect device data in real time and send the collected
data to iMaster NCE-FabricInsight — the DCN analysis component of Huawei iMaster NCE. Leveraging the intelligent fault
identification algorithm, iMaster NCE-FabricInsight can analyze network data, accurately display the real-time network status,
locate faults and identify their root causes in a timely and effective manner, and detect network problems that can affect user
experience, precisely guaranteeing user experience.
● CloudEngine 6855 series switches support insertion of IFIT extension headers into packets, path visualization, and interface-
level analysis of packet loss, traffic, and latency. This helps to achieve high-precision service-level packet loss detection and
facilitate fault demarcation.
● CloudEngine 6855 series switches support Packet Event. When a device discards packets due to reasons such as abnormal
forwarding, specified packet discarding rules, a full buffer, or ACL rule deny actions, or when the latency of packets exceeds a
specified threshold, the device reports related flow entries to the iMaster NCE-FabricInsight collector.
● CloudEngine 6855 series switches can interconnect with iMaster NCE-Fabric through standard protocols such as NETCONF
and SNMP to adapt to networks and implement automatic network management. This helps to provide more efficient and
intelligent operation methods, simplifying network management, and reducing the OPEX.
● CloudEngine 6855 series switches support the iLossless algorithm to eliminate packet loss on the conventional Ethernet. This
helps to build a lossless, low-latency, and high-throughput network environment for RoCEv2 traffic, meeting high performance
requirements of RoCEv2 applications.
● CloudEngine 6855 series switches support PFC deadlock prevention. The switches can identify service flows that may cause
PFC deadlocks and change queue priorities of these flows to prevent PFC deadlocks.
● CloudEngine 6855 series switches support Artificial Intelligence Explicit Congestion Notification (AI ECN). This future-oriented
function can intelligently adjust the ECN thresholds of lossless queues based on the live-network traffic model to ensure low
latency and high throughput with zero packet loss, maximizing the performance of lossless services.
● CloudEngine 6855 series switches support ECN Overlay. ECN Overlay applies ECN to a VXLAN network, enabling the traffic
receiver to detect congestion on the overlay network in a timely manner and instruct the traffic sender to reduce its packet
sending rate to relieve network congestion.
● CloudEngine 6855 series switches use a strict front-to-rear airflow design that isolates cold air channels from hot air
channels, meet heat dissipation requirements in DC equipment rooms.
● Air can flow from front to rear or from rear to front depending on the fan modules and power modules in use.
● Redundant power modules and fan modules can be configured to ensure service continuity.
● CloudEngine 6855 series switches use energy-saving chips and an intelligent fan speed control scheme to measure system
power consumption in real time. This can reduce O&M costs and help to build a green DC.
Clear indicators:
● The innovative port indicators can clearly show the port status, port speed, and status of all sub-interfaces.
● State and stack indicators on both the front and rear panels enable users to maintain the switch from either side.
● CloudEngine 6855 series switches support remote positioning. Users can turn on the remote positioning indicator through
the network management system (NMS) or console to easily identify the switches they want to maintain in an equipment
room full of devices.
Simple maintenance:
● The management port, fan modules, and power modules are on the front panel, which facilitates device maintenance.
● Data ports are located at the rear, facing servers. This facilitates cabling.
Basic functions
(including IPv6 and ● ● ●
VXLAN)
Telemetry ● ● ●
PTP ● ● ●
Multicast NAT ● ●
M-LAG virtual peer-
● ● ●
link
MACsec ●
AI ECN 2.0 ●
TCP optimization ● ●
NSLB
MoFRR ●
Adaptive routing
Automation ● ● ●
Basic intent
●
functions
Runbook ● ●
Multi-cloud and
multi-DC
●
automation
scenario package
Basic digital map
●
functions
Basic network
● ● ●
analysis functions
Network health
● ●
evaluation
Value-added
functions of
● ●
network traffic
analysis (100 VMs)
IFIT service
● ●
assurance
Value-added
functions of
●
network traffic
analysis (1000 VMs)
Value-added
package of multi-
●
cloud and multi-DC
analysis scenarios
Version mapping Select one from the three Used together with the Foundation, Advanced, or Premium package.
packages. The Advanced package
contains features of the
Foundation package, and the
Premium package contains the
features of the Advanced
package.
OpenFlow
Programmability
OPS programming
IPv4 dynamic routing protocols such as RIP, OSPF, IS-IS, and BGP
IP routing
IPv6 dynamic routing protocols such as RIPng, OSPFv3, IS-ISv6, and BGP4+
TCP6, IPv6 ping, IPv6 tracert, IPv6 socket, UDP6, and raw IPv6
Multicast IPv6 Layer 3 multicast and configuration of both Layer 2 and Layer 3 multicast services
LACP
BPDU protection
Traffic classification based on Layer 2 headers, Layer 3 protocols, and Layer 4 protocol
priorities
Traffic shaping
IEEE 1588v2
Telemetry
Enhanced ERSPAN
AI ECN
Intelligent lossless network
ECN Overlay
Enhanced NSLB
Configuration rollback
ZTP
MACsec
Microsegmentation
Command line authority control based on user levels, preventing unauthorized users
from using commands
RMON
Item Value
Note: This specification may vary between different scenarios. Please contact Huawei for details.
Hardware Specifications
Rated input voltage [V] 600 W AC&240 V DC power module: 100 V AC to 240 V AC, 50/60 Hz;
240 V DC
1200 W DC power module: –48 V DC to –60 V DC; +48 V DC
Input voltage range [V] 600 W AC&240 V DC power module: 90 V AC to 290 V AC, 45 Hz to 65
Hz; 190 V DC to 290 V DC
1200 W DC power module: -38.4 V DC to -72 V DC; +40 V DC to +57 V
DC
Typical power 138 W (100% traffic load, copper cables on 24 x 10G + 4 x 100G ports,
normal temperature, dual AC power modules)
156 W (100% traffic load, short-distance optical modules on 24 x 10G
+ 4 x 100G ports, normal temperature, dual AC power modules)
Number of fans 3
differential mode
DC power supply protection: 4 kV in common mode and 2 kV in
differential mode
MTTR (hour) 2
Availability 0.9999948505
AS/NZS: Australian/New Zealand Standard; VCCI: Voluntary Control Council for Interference
CE6855-48XS8CQ CE6855-48XS8CQ switch (48*10GE SFP+, 8*100GE QSFP28, Without Fan and Power Modules)
CE6855-48XS8CQ switch (48*10GE SFP+, 8*100GE QSFP28, 2*AC Power Modules, 4*Fans, Port-
CE6855-48XS8CQ-B
side Intake)
CE6855-48XS8CQ switch (48*10GE SFP+, 8*100GE QSFP28, 2*AC Power Modules, 4*Fans, Port-
CE6855-48XS8CQ-F
side Exhaust)
Fan module
Model Description
FAN-031B-F Fan box (F,FAN panel side intake), supporting the electronic label function
FAN-031B-B Fan box (B,FAN panel side exhaust), supporting the electronic label function
Power module
Model Description
PAC600S12-PF 600W AC Power Module (Front to Back, Power panel side intake)
PAC600S12-PB 600W AC Power Module (Back to Front, Power panel side exhaust)
PDC1K2S12-PB 1200W DC Power Module (Front to Back, Power panel side intake)
PDC1K2S12-CE 1200W DC Power Module (Back to Front, Power panel side exhaust)
N1-CE68AFRD2- SnS1Y N1-CloudEngine 6800 AI Fabric RDMA Application Acceleration Function 2-SnS-Year
N1-CE-F-LIC-MDCA N1-CloudEngine Data Center Switch Multi-cloud Multi-DC Value-added Package - Fixed
N1-CloudEngine Data Center Switch Multi-cloud Multi-DC Value-added Package, Per Fixed
N1-CEFMDCA - SnS1Y
device-SnS-Year
N1-CEAFP100VM-SnS1Y N1-CloudEngine Specified Flow Analysis Value-added Package Per 100 VM-SnS-Year
Application in a DC
On a typical DCN, CloudEngine 16800-X, 16800 or 8800 switches work as core switches, whereas CloudEngine 6855 series
switches work as TOR switches and connect to the core switches through 40GE, 100GE, or 200GE ports to build an end-to-
end and fully-connected 100GE/200GE/400GE network. The switches use VXLAN and other fabric technologies to establish a
non-blocking large Layer 2 network, which allows large-scale VM migrations and flexible service deployments.
Notice
The purchased products, services and features are stipulated by the contract made between Huawei and the
customer. All or part of the products, services and features described in this document may not be within the
purchase scope or the usage scope. Unless otherwise specified in the contract, all statements, information, and
recommendations in this document are provided "AS IS" without warranties, guarantees or representations of any
kind, either express or implied. Huawei Technologies Co., Ltd. Address:
Huawei Industrial Base Bantian,Longgang
The information in this document is subject to change without notice. Every effort has been made in the
Shenzhen 518129 People's Republic of China
preparation of this document to ensure accuracy of the contents, but all statements, information, and
recommendations in this document do not constitute a warranty of any kind, express or implied. Website: www.huawei.com