Expedition

Log Analysis Feature Guide

Version 1.0.5

1
Palo Alto Networks, Inc.
www.paloaltonetworks.com
© 2020 Palo Alto Networks, Inc. Palo Alto Networks is a registered trademark of Palo Alto Networks. You can find a list of our
trademarks at https://www.paloaltonetworks.com/company/trademarks.html. All other marks mentioned herein may be
trademarks of their respective companies.
Revision Date: Apr 11, 2022
Contents

What are the Log Analysis Features in Expedition? 4

Which features should you choose? 4

Steps for Expedition Log Analysis Features 6

Adding a PAN-OS device in the device tab 7
Adding a Next-gen firewall and retrieve it’s contents 7
Adding a Panorama and retrieve it’s contents 11
Creating a new Expedition project 12
Project Settings 13
Manage Devices 13
Adding a PAN-OS device as the Log Connector 14
Importing the latest running configuration from PAN-OS device (Panorama or firewall) 15
Selecting Security Policies 16

APP-ID Adoption 17
Retrieve APP-IDs 17
Split Rules Known/Unknown 19
Clone the existing rule 20
APP-ID Reconciliation 21

Common Steps for Machine Learning (ML) and Rule Enrichment (RE) 23
Machine Learning (ML) Settings 23
Importing Logs to Expedition 24
Exporting from NGFW 24
Manually Export Logs From MONITOR 26
Expedition as Syslog Server 26
Importing from Splunk 27
Configure M. LEARNING on the device 29
Processing Logs From Selected Devices 30
Debugging Log Processing 30

Rule Enrichment(RE) 31
Enable Rule Enrichment 31
Rule Enrichment Discovery 31
Import Discovered Rules back to Project 33
Debug the Rule Enrichment 34

MACHINE LEARNING (ML) 35

Enable Machine Learning 35
Machine Learning Discovery 36
Discovered Servers 39
Import suggested rules back to Project 41
Debug the Machine Learning 43

© 2020 Palo Alto Networks, Inc. 2

Push Modified Security Policies back to PAN-OS Devices 44
Generate API Requests 44
Send API Requests 45

Revision History 47

© 2020 Palo Alto Networks, Inc. 3

What are the Log Analysis Features in Expedition?

The Expedition provided below log analysis features that can help you to either migrate your service-port based
security policies to APP-ID based security policies or to refine your existing security policies based on traffic
logs.

● APP-ID Adoption: This feature is used when you have a set of service-port based security policies and
you would like to migrate those legacy security policies to APP-ID based security policies. Expedition will
retrieve APP-ID info from the active log connector, this feature does not require Firewall logs stored in
Expedition.

● Rule Enrichment (RE): This feature is used when you would like to tighten your existing security policies
to remove “any” in security policies. For example, you have a security policy that contains “any” in either
applications, users, zones, or services fields. This feature will auto discover APP-ID and service port info in
the firewall traffic logs to see if it matched the application-default ports and help you to tighten your
security policies to replace “any '' with correct APP-ID and service ports in the security policy. Expedition
will process firewall log in csv format , so this feature does require Firewall logs stored in Expedition.

● Machine Learning (ML): This feature will suggest new sets of security policies based on analysis of the
firewall traffic logs. It is often used in Greenfield deployment or when you have a set of rules that’s more
permissive than required and you don’t know what security policies are required. The ML process will
identify servers, consumers and provide all the security policies including source, destinations, APP-ID,
and service-ports . Expedition will process firewall log in csv format , so this feature does require Firewall
logs stored in Expedition.

Which features should you choose?

Depending on your use case , you will choose different log analysis features. Below are some use cases
and what features you can use.

● Use Case #1

I am clear on what security policies are required in my environment. I want to migrate all my
service-port based ( Layer 4 )security policies to APP-ID (Layer 7) based security policies.

Feature: If you are clear on the security policy requirements, you will choose APP-ID Adoption, with
this feature, you do not need to store firewall logs in Expedition. (if the PAN-OS device is running
PANOS 9.1 or later, it is recommended to directly use Policy Optimizer feature in PAN-OS)

● Use Case #2
I want to migrate all my service-port based ( Layer 4 )security policies to APP-ID (Layer 7) based
security policies and would like to see if Expedition can auto suggest new APP-ID based rules based on
live traffic logs.

© 2020 Palo Alto Networks, Inc. 4

 Feature: If you are looking for security rules suggestions based on live traffic logs, you will choose RE or
ML , however these features require you to store firewall logs in Expedition. If you can’t send traffic logs
to Expedition, you will choose “APP-ID adoption” instead.

● Use Case #3

I am not clear on what security policies are required in my environment , also I want to migrate all my
service-port based ( Layer 4 )security policies to APP-ID (Layer 7) based security policies.

Feature: Machine Learning (ML). This feature will help you by suggesting a new set of security policies
per APP-ID with detailed source and destination info.

● Use Case #4

Most of my security policies are APP-ID based security policies and some of them contain “Any” in the
service field, I would like to see if I can tighten the security policies to only use “application-default” or
the ports that’s needed.

Feature: Rules Enrichment (RE). This feature will help you refine the existing APP-ID based policy.

● Use Case #5

We have security policies that are more permissive than what we really required and I would like to
know what security policies are actually required.

Feature: Machine Learning (ML). This feature will help you by suggesting a new set of security policies
per APP-ID with detailed source and destination info.

● Use Case #6

This is a greenfield firewall deployment and we are not clear on what security policies are required ,
most of the security policies are too permissive.

Feature: Machine Learning (ML). This feature will help you by suggesting a new set of security policies
per APP-ID with detailed source and destination info.

© 2020 Palo Alto Networks, Inc. 5

Steps for Expedition Log Analysis Features
Here is the list of the steps for the log analysis features:

1. Adding a PAN-OS device in the device tab and retrieve latest contents
2. Creating a new Expedition project
3. Adding Panorama or firewall as Log-Connector
4. Importing the latest running configuration from PAN-OS device (Panorama or firewall)
5. Selecting Security Policies
6. Required Steps for ML and RE (For APP-ID Adoption, please jump to step 7)
a. Configure Machine Learning settings
b. Importing traffic logs to Expedition
c. Configure M.Learning on the device
d. Processing the logs from the selected PAN-OS device
7. Please refer below individual sections for the different subsequent steps
a. APP-ID Adoption
b. Rule Enrichment (RE)
c. Machine Learning (ML)
8. Push Modified Security Policies back to PAN-OS Devices

© 2020 Palo Alto Networks, Inc. 6

Adding a PAN-OS device in the device tab

Please follow the steps below to add the PAN-OS device in the device tab and retrieve the latest contents form
the device.

Adding a Next-gen firewall and retrieve its contents

Expedition supports all the PanOS versions since version 7.0 up to 10.x. Let’s follow an example on how to create
a new Device and import the configuration and securely store it on Expedition.

a) Navigate to the DEVICES tab

b) Add a new Device by clicking on the plus button located on the top-right from the panel.

c) A new window will be shown to fill with all the information required.
● Device Name: It’s the name you want to call your firewall
● Model: Palo Alto Networks device model
● Hostname/IP: IP or name used to connect to your firewall, if it’s a name Expedition needs to know how
to resolve it, check the DNS used by Expedition it’s the right one. You can check from the CLI
# sudo cat /etc/resolv.conf
● Port: where the management is running, by default 443
● Serial #: This field is required and will be used as an Index to use the right one.
● Serial # HA: In case this firewall is part of a Cluster you can set the HA serial. This will matter for the
Machine Learning module which will be explained in another chapter of this document.

● Click on Save to add the Device to Expedition

© 2020 Palo Alto Networks, Inc. 7
 d) Once the Device has been created and listed from the Devices view we have to edit and add the credentials to
retrieve the contents like applications database, system information and the configuration. Select the device
and double click to edit it or by clicking on the Edit button (pencil).

e) The Edit Device window now is displayed. From the Configuration Tab let’s add our credentials to connect to
the firewall and Expedition will request the firewall to generate a new API key.
notice the generated API Key will be valid as long as the user doesn't change the password from the firewall.

● Click on the plus icon to add a new API Keys

● Auth. Type: How we want to authenticate against he firewalls, we can choose to provide
username and password and let Expedition request the API key to your firewall or in case you
already have the API key choose API KEY and paste your key in the text field

© 2020 Palo Alto Networks, Inc. 8

In this example we are going to use Username and Password and provide them:
● Role and Apply all Roles: When you add a new API Key this can be attached to a Role inside
Expedition, that means when you have a user from Expedition with Role admin inside one Project
and that user tries to push changes using API Keys Expedition will use the API Key based on the
user’s Role in this example admin. If you didn’t add an API key to the admin role that user will be
unable to send any API Call out.

For small environments where you will have only one user and it will be admin there is no need to check
the Apply all Roles and keep that key only attached to the admin Role.
● Click on the Add blue button to generate the Keys.

● Navigate to CONTENTS to retrieve the Running configuration.

© 2020 Palo Alto Networks, Inc. 9

 ● Once the download process is completed, you will see the “State” column changed to
“Downloaded” , then click on “Save” .
Expedition downloaded the device configuration and stored it in the hard-drive encrypted. You can check the file
from the CLI by entering in the following folder:
$ cd /home/userSpace
$ cd devices
$ cd <the device serial number >
$ ls -la

For debugging purposes there is a file on the devices folder called “debug.txt”. The content of this file comes from
the daemon who controls the access to the firewalls so expect to find the API requests made to retrieve the keys
or dynamic reports in some cases, for security we stripped out the parameter key from the request. So, in case you
want to re-use an API call you will have to add at the end the &key=<and your API key>

© 2020 Palo Alto Networks, Inc. 10

At this moment Expedition keeps a snapshot of your running configuration. In case you make any change on the
device and you want to update your snapshot you have to edit the Device again and retrieve the running
configuration again.
If you were already using that config on a project and you want to import the changes there is no other way than
remove the current config from the project and import the device config again, all changes you made on that
config and not exported will be lost.

Adding a Panorama and retrieve it’s contents

Importing a Panorama is really similar to import a Next-gen Firewall and here we will show only what’s is different

You can do the same steps used to create a new Next-gen Firewall, generate the API keys and retrieve the
Contents, now a new tab called PANORAMA DEVICES will be active, so click on that tab.

1. If we need to play with the firewalls connected to that Panorama we need to know them first, so click on
Retrieve Connected Devices. This will request that information to Panorama and create the devices on
Expedition but referencing them under this Panorama device. That means all the requests we will
generate to talk with that Firewall will be done using Panorama as a Proxy.

2. In case we need to download the configuration of one of these devices just select it from the list and click
on Retrieve Contents. This will allow you to import the configuration of that firewall into your project
once we create the project.

If you have one firewall created in Expedition and then you import Panorama and the that firewall it's listed as one
of the connected devices to that panorama (Serial Number is used as index here) the existing firewall in
Expedition will go under the Panorama management meaning from that moment the API requests against it will
pass through Panorama.

From the Devices view we can hide all the connected devices to its Panorama by using the following buttons from
the Panel header

© 2020 Palo Alto Networks, Inc. 11

Creating a new Expedition project

In Expedition each project has its own database. You can create as many projects as you want.

Let’s create a new project and see the workflow we have to follow:

a) Click on the plus button to create a new Project

b) A new window will pop up, assign a name to your Project and in case you already created a Device you can
select it from the combo box, when you select a firewall or panorama on that combo box you are forcing
Expedition to import the same Applications database to your project, that Applications database was
downloaded to Expedition at the same time we were retrieving the configuration. Doing this your Project
will have the same applications that your Firewall.
c) Purpose if this Project: this is used to provide some statistics but doesn’t affect in nothing
d) Click on Create Project.

In case we selected a device, this will automatically be added to your project, you can edit these settings to
manage users and devices from within the Project.

© 2020 Palo Alto Networks, Inc. 12

Project Settings
After the project creation we can continue managing the project , we can add more users or devices to the project
and import or export the project to be shared with other Expedition instances.
To Edit the Settings just select the Project and click on Settings

Manage Devices
If we want to allow access to import a configuration from an existing device or enable the users to generate API
calls to be sent to the device you must allow the project to have access to that device.
To do it just click on Devices and select the firewalls you want to allow and move them to the right panel

Enter into the Project and Import the configuration from the device who has the configuration (Panorama or a
firewall) and remember that device must be in your Allowed Devices list.

© 2020 Palo Alto Networks, Inc. 13

Adding a PAN-OS device as the Log Connector
The log connector is used by Expedition like a FILTER to know from all the information is stored as Parquet files
from what serial numbers and vsys we must dig in only and skip other information coming from a different serial
and vsys.

Under the plugins tab, add a log connector. Give it a name, select the device and the device group, the firewall, and
choose the time period for log analysis. This view will show you Device-Groups or Vsys based on the configuration
you were importing as a Base configuration in your project. The time period can be custom as well.

Ensure the Log connector is Active.

© 2020 Palo Alto Networks, Inc. 14

Importing the latest running configuration from PAN-OS device (Panorama or firewall)
In the Projects tab, Double Click on the project you created in the previous step, go to “Import” and double click
on the device to import the running configuration into the project.

© 2020 Palo Alto Networks, Inc. 15

Selecting Security Policies
Navigate to “POLICIES”-> “Security” , you will select from the right lower bottom drop down menu for the vsys
or device group policy you would like to work on:

Once you select the correct vsys or device group, you will see the background color of policies turn from gray to
white , that means you can now edit the policies, in below example, I have a firewall configuration and the security
policy is in vsys1 , so I will select vsys1.

© 2020 Palo Alto Networks, Inc. 16

APP-ID Adoption
Expedition APP-ID Adoption feature can help you to convert port-based rules to application-based rules enables
you to include the applications you want to allow in an allow list and deny access to all other applications, which
improves your security posture. Restricting application traffic to its default ports prevents evasive applications
from running on non-standard ports. Removing unused applications from rules is a best practice that reduces the
attack surface and keeps the rulebase clean.

❏ Retrieve APP-IDs
❏ Split Known/Unknown Rules
❏ Clone the existing rule
❏ APP-ID Reconciliation

Retrieve APP-IDs
Go to “POLICIES” -> “Security” , select the Port-Based security policies , and right click “Retrieve APPs(Fast)” or
“Retrieve APPs(Slow)” -> “Selection” , alternatively , you could also choose “All Rules” if most of your rules are
port-based.

After the Retrieving APPs process finished , you will see a new column “APP-ID via Log” shown below and it listed
all the APP-IDs found in the firewall traffic logs . Below is the result when you select “Retrieve Apps (Fast)” , the
report will be generated faster without traffic size detailed, for an environment that has a lot of traffic logs, please
select this option for faster processing.

© 2020 Palo Alto Networks, Inc. 17

The difference between “Retrieve APPs(Fast)” and “Retrieve APPs(Slow)” are the “Retrieve APPs(Slow)” will
take longer time to process and show you the detail traffic size based on APP-ID as shown in below screenshot:

© 2020 Palo Alto Networks, Inc. 18

Split Rules Known/Unknown
When discovered APP-ID contains Unknown-TCP or Unknown-UDP, it’s better to split them to different rules to
further analyze the traffic logs later. To split the Known rules from Unknown rules , right click on the rules , select
“APP-ID Adoption” -> “Split Rules Known/Unknown” -> “Selection”.

After you performed the step, You will see the original security policy got split into two rules , original rule will
contain known APP-IDs in the “APP-ID via Log” field , Expedition cloned another rule on top of the original rule
and add a prefix “Unk-” to the original rule name , the example below, the original rule name is “AllowAllOut”, the
“Unk-AllowAllOut” contains the “Unknown-TCP” and “Unknown-UDP” as shown in below screenshot:

© 2020 Palo Alto Networks, Inc. 19

Clone the existing rule
Right click on the rule that has discovered apps , select “Rule Actions” -> “Cloned Selected Rule” -> “Below”.

This action will clone the existing rule and put it under the existing rule like below screenshot with prefix “CL-”, the
example below, the original rule name is “AllowAllOut”, the new cloned rule will be named “CL-AllowAllOut”. This
is identical to the original service-port based rule. The purpose of this step is to keep the original rule in the
bottom of the APP-ID based rule.

© 2020 Palo Alto Networks, Inc. 20

APP-ID Reconciliation
We will highlight the original security policy , right click to select “APP-ID Adoption” -> “APP-ID Reconciliation”
-> “Applications” -> “Selection” . This step is to add the app-ID that's in the “APP-ID via Log “ field to the
“Application” field of the rule, so it is converting the existing service-port based rule to “APP-ID” based rule. You
have three options in this process:

1. Application - The application field will be replaced with APP-IDs seen in the “APP-ID via Log” field.

2. Recommended Application- The application field will be replaced with recommended APP-IDs. For
instance if the APP-ID seen in the traffic log is “slack-base”, the recommended APP-ID will be “slack”
which is a parent APP-ID that covers “slack-base” and other APP-IDs under the same parent.

3. Recommended Application & Dependencies - - The application field will be replaced with recommended
APP-ID and it’s dependencies , for example : if “gmail-chat” is seen in the “APP-ID via log” field , and it will
add it’s parent application “gmail” also the dependencies , “google-base” to the rule as well. You can find
APP-ID dependencies by looking up the Application database in the PAN-OS device.

Recommended Application:

© 2020 Palo Alto Networks, Inc. 21

Recommended Application and dependencies in PAN-OS application:

After the step is performed , you will see the original service-port based rule become a APP-ID based rule as
shown in below screenshot, and the APP-IDs shown in APP-ID via Log field will be move to “Application” field of
the rule , the “insufficient-data” won’t be move to Application field , this is seen when Firewall does not have
enough packets to determine the APP-ID.

© 2020 Palo Alto Networks, Inc. 22

Common Steps for Machine Learning (ML) and Rule Enrichment (RE)
❏ Configure Machine Learning settings
❏ Importing traffic logs to Expedition
❏ Configure M.Learning on the device
❏ Processing the logs from the selected PAN-OS device

Machine Learning (ML) Settings

As we described in the overview, we can set up Expedition to run standalone or splitting in GUI and Analysis, that
is only the case if we have a shared Expedition VM with more hardware assigned to run faster analysis.

To configure it and specify where the Parquet files will be stored after the log process we have to navigate from
the Dashboard to the SETTINGS tab, then select the M.LEARNING tab.

Step 1: Check the IP address shown in Expedition ML Address is your own IP address if you are running in
standalone mode. If not, put the IP address of the other Expedition VM instance with more hardware resources
than the one you are currently configuring.

Every time you enter here if the ip address is 127.0.0.1 Expedition will replace it by the one shown in the browser’s URL.

Step 2: Type where you want to store the data after the process in parquet files. You could specify a different path
to store your firewall traffic logs. If you don’t have any folder created in Expedition let’s create one, in this case we
will use one named PALogs.

sudo mkdir /PALogs

sudo chown -R www-data.www-data /PALogs

with these commands we are creating and allowing “apache” to have write access to it.

Step 3: click on the Save button located at the bottom bar – right to activate the changes.

If at some point we want to start fresh and remove all the data processed as parquet files we can click on the red
button “DELETE ALL DATA STRUCTURED FILES”. That process cannot be undone.

© 2020 Palo Alto Networks, Inc. 23

Importing Logs to Expedition
There are below four ways to import traffic logs to Expedition:

1. Exporting from NGFW

2. Manually Export Logs From MONITOR
3. Expedition as Syslog server
4. Importing from Splunk

Depending on your platform and or your log storage strategy, you may want to choose between those four
methods. For instance, you may have a large amount of firewalls deployed on prem and on cloud environments
that you would like to analyze and that are managed by a Panorama device. In such a case, probably the best
strategy would be to use Expedition as a Syslog server and request the firewalls to submit the traffic logs to
Expedition via a Syslog log forwarding profile.
Maybe you already are having a Syslog log forwarding profile on your security rules to store the traffic logs in a
Splunk instance. In that case, maybe the best approach would be to select the fourth option.

Below we describe the four alternatives and discuss the benefits of using each one of them.

Exporting from NGFW

There is a functionality in Palo Alto Networks Firewalls that allows you to automatically export the logs on a daily
basis and import it automatically on Expedition, that functionality is called “Scheduled log export”.

Note: The PA-7000 Series and Panorama devices do not offer such log Export functionality or it is limited to the first
1.000.000 entries

https://docs.paloaltonetworks.com/pan-os/10-0/pan-os-web-interface-help/device/device-scheduled-log-export.
html

Actions
We will configure the Scheduled Log Export with the following settings.
● The log type to select is “traffic”;
● we will set the Hostname with our Expedition IP or hostname (if resolved by your DNS server);
● the start time could become something like “00:15” or a time that you consider suitable for initiating a
SFTP transfer that could be between few MBs up to a couple or TBs (depending on your infrastructure
traffic);
● we will set the destination folder in which traffic logs will be stored in Expedition. We recommend using
the folder /PALogs, which may already exist in your Expedition instance, and allows the users “expedition”
and “www-data” to access it;
● we will set the username as “expedition”, which it is an Expedition cli user that we have by default created
in the Expedition VMs and that has privileges to store files in required file system spaces; and
● we will set up the password as “paloalto” or the defined password that you have for the expedition user in
your Expedition VM.

When you create this for the first time you should run the “Test SCP Server connection” twice. The first time in
order to exchange the SSH keys and the second to write a test file and validate that the PanOS Firewall can write
in the folder provided under “path”.

© 2020 Palo Alto Networks, Inc. 24

As this folder could need a lot of storage in the beginning it is a good practice to add a new VM DISK and mount it
under /PALogs. Be sure Expedition can write if you chose to use expedition as the user to write the data. This can
be done with the following command:

chown -R expedition.www-data /PALogs

Benefits
On the positive side, this action is performed only once a day and can be programmed to execute during the night
or other low load period of the day.
Additionally, we can be confident that all the traffic logs will be submitted to our Expedition if there is enough
space in our Expedition VM to host the file, as the transfer is performed via a SFTP connection.
You do not need to have the VM running all the time to receive the traffic log files. If you have set up the
scheduled time to be at 10:00am, this gives you the chance to have the Expedition VM off during night hours and
bring it up again before the SCP is going to take place.

Drawbacks
This approach requires firewall configuration changes, in order to set up a periodic log export.

Manually Export Logs From MONITOR

You can always go into any firewall from Palo Alto Networks and from the Monitor tab export the logs in CSV
format and upload that CSV file to Expedition for processing, concretely, into your “PALogs” folder.
The format of those CSV files would be the same as executing a Scheduled Log Export.

Actions
Log into your PANOS device and, under the Monitor tab, enter the filter that you would like to use for collecting
traffic log entries. For instance, specify the rule name and destinations that you would like to study.

© 2020 Palo Alto Networks, Inc. 25

Afterwards, click on the “Export to CSV” button, which resembles an Excel icon, to download the filtered traffic
logs into your computer.
Afterwards, upload the obtained CSV into the /PALogs folder in your Expedition by using the SFTP client of your
choice.

Benefits
This option offers us an opportunity to collect specific logs from the firewall, selecting a period of time, specific
sources, destinations, and/or rule names that we would like to bring into Expedition and further study.

Drawbacks
However, this approach would not be suitable for a periodic task, as it requires manual connection to the Firewall,
set the filter logic we want to apply, export the log in a CSV format and manually upload it into the “/PALogs” folder

Expedition as Syslog Server

In case the scheduled approach would not fit with your requirements, either because you are not interested in
secure copying all the traffic log entries into Expedition, or because you are dealing with a platform that limits to 1
million the number of entries to be exported, then you may prefer using a syslog message approach.

Expedition can become a Syslog server that can receive the logs generated by platforms like the PA7000 or in any
case you want. And we can submit traffic log entries to our Expedition using a log forwarding profile.

Actions
Become root in your Expedition instance to navigate to the following folder.

sudo su
cd /var/www/html/OS/rsyslog

In that folder you will find 3 examples for the rsyslog.conf you can use:
● rsyslog.default-tcp,
● rsyslog.default-tcpudp and
● rsyslog.default-udp

Each one listen in the specified protocol and port and allow Expedition to accept syslog from some specified
networks (in the examples we define 127.0.0.1, 10.11.29.0/24, 172.16.26.0/24, *.paloaltonetworks.com) and
store them in “/PALogs”.
Restart the VM to make sure that all required syslog modules are activated. Restarting the rsyslog service may
not be enough as dependent modules may not have been initialized with your required settings.

Benefits
Using Expedition as a syslog server may help you determine the security rules that you are interested in studying,
as we can determine the rules that are going to have the log forwarding profile applied.

Drawbacks
This option requires you to activate Expedition as a syslog server and to tune the rsyslog settings to allow
connections via TCP and/or UDP from a specific set of IP addresses.
This option requires your Expedition instance to be available 24x7, as traffic logs are submitted to your
Expedition in real-time. If the syslog entries are defined to be submitted using UDP protocol, there are higher
chances that some traffic logs may be lost if the connection with Expedition is interrupted.

© 2020 Palo Alto Networks, Inc. 26

If you already have log forwarding profiles applied to your security policy rules, you may have to generate
different combinations of log forwarding profiles. Rules can only have one log forwarding profile applied, even this
profile can combine multiple syslog servers. Therefore, you may have to create a new log forwarding profile that
contains your prior server as well as the Expedition syslog server, and apply this new log forwarding profile into
the rules of your choice by replacing the prior profile applied.

Importing from Splunk

Finally, there is a chance that you have already log forwarding profiles active in your devices to submit the traffic
logs to an Splunk instance. We have prepared Expedition to be able to collect traffic logs information for Splunk,
generating the API queries that would be required to access those entries.
As we do not require the whole amount of information that traffic logs contain, the queries that we generate to
interact with Splunk have already been constructed to request only the needed log fields and to create
aggregations already in Splunk in order to reduce the amount of data that needs to be submitted from Splunk to
Expedition.

Actions
First, make sure that your Splunk instance has the “Palo Alto Networks App for Splunk” and “Palo Alto Networks
Add-On” installed, as this will provide the required schema to query Palo Alto Networks logs. For more details on
the apps, please refer to the link: https://splunk.paloaltonetworks.com/

From Expedition UI, select the device that you would like to retrieve traffic logs from. On the M.Learning tab, click
on the “Query Logs from Splunk®” button to open the Splunk query window.

Fill in the form to generate the individual queries that will be sent to the Splunk instance. In the figure below, we
have provided the IP address of our Splunk instance using the default’s Splunk port 8089 and using the “admin”
account that has API permits. We have defined the period of time we want to collect data from, in the example,
between the first and the third of November 2020 and using the “paloalto” index to increase the performance on
Splunk while loading the log files that refer to Palo Alto Networks devices.

To generate the queries, click on the “Prepare Queries” button. This will create one query per day and those will
be presented in the bottom grid. At this point, we can fire the queries to Splunk by clicking on the “Retrieve Traffic
Logs” button.

At this point, the queries will be submitted to splunk and Expedition will be monitoring the query status to
download the results once Splunk has reported the query as completed. Depending on your Splunk settings and
the number of queries submitted, some of them may be in a queue.
Once the query has been completed in Splunk, and Expedition has downloaded the results, you will be presented
with the message “File successfully downloaded”, and you should be able to display the file in the prior M.Learning
view.

© 2020 Palo Alto Networks, Inc. 27

Benefits
There is no need to interact with your Palo Alto Networks devices, as those are already submitting the traffic logs
to Splunk. instead, the interactions are performed via API with your Splunk instance.
If you have indexes created, requests can make use of them to speed up the process.
We can specify the periods of time that we want to study, and only retrieve the traffic logs that were generated
during those dates.

Drawbacks
It requires a user account with API querying permission in order to interact with Splunk.
Currently, this approach offers manual query generation and execution and it needs to be executed for every
firewall that we want to collect traffic logs from.
In the future, we will provide opportunities to retrieve Device Groups traffic logs directly from the Panorama
point of view.

© 2020 Palo Alto Networks, Inc. 28

Configure M. LEARNING on the device
Once the device has been added in the previous step, we have to edit it and go to the M.Learning Tab.

From here we will assign the PATH where our logs have been imported in Expedition, or via SCP or SYSLOG, then
click on Search Files to allow Expedition to search on it and show us all the related logs to the device we are
configuring.

Expedition will filter out and show you only the files from the configured folder (and recursively from there) only if
the serial number configured on the device (or the HA Serial) matches the serial seen in the csv files. So, if you
configure the device and you set the serial “123” but the CSV content talks about the serial number “567” ; those
logs won’t be shown under the view.

By default, the first time Expedition sees a new file is flagged as a NEW file. When clicking on Process files all the
NEW files will be converted to Parquet.

We can select an action for “After Process” like we want to delete the files after transformation to parquet to
reduce the space we need to store the raw data.

If we want to avoid processing some file we have to just select it and click the icon (Ignore File), that will tell
Expedition to skip that file at the Process time.

In case you are using SYSLOG to receive logs in Expedition check the “Log Files come from Syslog” this will
prevent you to transform the logs from Today since it’s a file is still opened and used by the syslog server before
the log rotation that will occur when we change the day at 00:00, so this check will hide the log from Today until
Tomorrow.

© 2020 Palo Alto Networks, Inc. 29

Processing Logs From Selected Devices
We can enter on each firewall where we are importing more files for the process and click on “Process Logs” or we
can do it from the Devices view by selecting the Device and clicking on the blue Icon.

Debugging Log Processing

While Expedition is processing the logs, we can view the progress and potential issues from the CLI by doing this:

tail -f /tmp/error_logCoCo

© 2020 Palo Alto Networks, Inc. 30

Rule Enrichment(RE)

As we described before RE it's useful to reduce the surface attack in your current security policies, today with RE
we can remove all the “any” from our Rules. One of the most popular use cases is to help to enrich a policy based
on services with App-ID for instance.

❏ Enable Rule Enrichment

❏ Rule Enrichment Discovery
❏ Import Discovered Rules back to Project

Enable Rule Enrichment

In this example we have Rule Enrichment Discovery

After tag the Rule click on DISCOVERY green bottom located in the bottom bar (center)

A new window will pop up. Click the TAB called RULE ENRICHMENT. Then you have the option to override the
Time Range defined previously in the LOG CONNECTOR, if you define like the last 30 days in the connector you
can select a specific time frame like example and that will take preference from the connector configuration. Also

© 2020 Palo Alto Networks, Inc. 31

you can specify the thresholds to discard traffic when it's less than specific bytes or hits. Then click on “Analyze
Data” and wait

Once the analysis has finished we will get something like this, this information is based on the logs previously
processed and stored as parquet and filtered by the log connector and the name of the rule tagged as RE (you can
tag as much rules as you want and perform the analysis at the same time for all the rules, each one will create its
own results)

© 2020 Palo Alto Networks, Inc. 32

From the output we can read there is only one Source IP address, I can read the applications (dns) and Expedition
has calculated if the port seen using the applications is the same as defined as application-default to tell us the
service can be considered “application-default” if cannot be identified you will see the port discovered (in this
example should be udp/43)
The major difference between RE and ML is RE produced less rules than ML, it will only produce 4 -6 rules based
on Application and service ports , the rule will be group to below 6 categories based on below:
1. application is Unknown-udp
2. application is Unknown-tcp
3. application is insufficient data
4. application with application-default port
5. application with custom service port.
6. application is incomplete

© 2020 Palo Alto Networks, Inc. 33

Import Discovered Rules back to Project

To import the information to our Rule we have to click on the option “IMPORT INTO PROJECT” under the
“Analyze Data” button.
I will select Applications, Service and Source, for the Source I will check IPs that means bring the IP address, I can
choose the regions found instead.

We can override the current (original) rule with the new data to improve that rule or we can let Expedition clone
the existing rule and import the new data on the new cloned rule, that cloned rule will be stored on top of the
original one. This will allow you to run more analysis in the future to ensure you don’t miss anything because the
period of time analyzed was not enough to capture all the data.
Click on “Import”.

How we left “Replace Rule” instead of cloning, Expedition will automatically Clone the Original Rule, bring the
new data on top and disable the current Rule for you to validate that everything looks good after the enrichment,
you can remove the disabled rule at the end of the process.

When we use Rule Enrichment Expedition will group all the data by Rule Name and it will create a rule with all
the traffic seen with Users, another one with the traffic seen without users, another one with the traffic where
the applications where found through their default-port and another one with the traffic where the applications
were found on a port different the default port that means with Rule Enrichment we will get a maximum of 4
rules by each rule we are analyzing.

© 2020 Palo Alto Networks, Inc. 34

Debug the Rule Enrichment

tail -f /tmp/error_SecRulesEnrich

on the red button “DELETE ALL DATA STRUCTURED FILES”. That process cannot be undone.

© 2020 Palo Alto Networks, Inc. 35

MACHINE LEARNING (ML)

Machine Learning is a more sophisticated process than RE. ML was designed to help companies to know what is
going on in the network and translate that to security policies based from the beginning on App-ID and User-ID.
The most common case we cover is a “Greenfield” where you have one rule covering all the traffic and Expedition
suggests all the new rules that cover the traffic seen.
In the ML process Expedition identifies the Servers in the networks based on some analysis since we support
environments with asymmetric traffic this analysis is critical at the time to suggest new rules. Here are the
specific steps for ML:

❏ Enable Machine Learning

❏ Machine Learning Discovery
❏ Import Suggested Rules back to Project

Enable Machine Learning

Example like below , we want to replace below one rule that is too wide open since it allows everything between
two networks.
We want to know what is exactly happening and create the rules based on what we have learned.

First step is to enable the “Machine Learning feature “ by highlighting the rule and right-clicking to select
“Machine Learning” -> “Monitor” -> “Selection” .

© 2020 Palo Alto Networks, Inc. 36

Machine Learning Discovery

After tag the Rule click on DISCOVERY green bottom located in the bottom bar (center) , select “Machine
Learning”

© 2020 Palo Alto Networks, Inc. 37

A new window will show up and we can override the time to analyze overriding the one configured in the LOG
CONNECTOR (remember to check you have a LOG CONNECTOR configured and Active). Also you can select to
discard traffic if it’s less than bytes or less than certain hit counts you specified in the threshold.

© 2020 Palo Alto Networks, Inc. 38

The “Analyze Data” button can be extended by clicking on the arrow.

● Cloud: Means Expedition will consider some applications as Cloud, when found in the traffic the
destination ip addresses will be considered as “any” since they can dynamically change and doesn’t make
sense to keep the ones the network resolved in the moment we captured the traffic.
● Common: Are considered common applications that are present in all the networks and generate a huge
volume of logs. Ex: ping, dns, ldap. In some cases you won't want to waste resources to analyze logs
related to those applications to speed up the analysis for other applications. In case Expedition finds
traffic regarding those applications will be considered as source “any '' and destination “any ''. Ex: ping
Rule suggested ANY – ANY – ping – ALLOW
● Peer-to-Peer: All the applications classified as peer to peer by Palo Alto Networks.
● Global: All the other applications. Expedition will analyze sources, destinations, users, service ports to
suggest Rules based on how they are consumed.
Once you select the Application categories, you can then click on “Analyze Data”
Once the analysis has finished:

© 2020 Palo Alto Networks, Inc. 39

This is the result of the analysis for a single rule. This is all the Rules suggested based on how the applications
have been consumed. In this case there are no users. The ML result will generate Rules based on Application , so
for example if you have multiple sources and destinations with “SSL” as application , it will generate multiple
security rules based on different sources and destinations, in contrast to RE feature, RE feature will group all
different sources and destinations to one rule. The Flow has been calculated after figuring out who the servers
are on the networks. Tag tells you the container for the APP. In this case all were Global.

Discovered Servers

This is the list of the servers we found from the logs after analyzing who is who in the network. This is
important to Expedition to understand the flow of the communications.

This list of servers can be exported for an offline review by clicking on the Export Excel.

We can group the discovered Servers by Applications and see how many Applications each Server has.

● Point your mouse over the column “Server IP Address” and click on the arrow and click on the “Group
by this field” In this example nothing will change but in real life environments will give you the
applications served by each server we used.

© 2020 Palo Alto Networks, Inc. 40

Import suggested rules back to Project
Review the new suggested rules and select the ones you want to import to your security policy

© 2020 Palo Alto Networks, Inc. 41

In our example we will select them all and we will import everything, you can be more selective and convert all the
new objects as part of your shared objects instead of creating them under the vsys / device-group selected at the
time to import. Remember you can always export as Excel and review them offline.
In some cases , you would like to implement the security policy based on source or destination subnet instead of
single IP, you can click on “Advanced Options”

It will then switch to the advanced option window like below, you could specify you want to import source or
destination as /24 subnet when you see more than “x” IPs in the same /24 subnet. In the example, we use “2” , if
the suggested rules contains more than 2 different IPs (10.0.0.1 and 10.0.0.2) in the same subnet , the import will
change the source to “10.0.0.0/24” subnet instead of import them as 2 single IPs.

© 2020 Palo Alto Networks, Inc. 42

After clicking on “Import” , the rules imported will be named as “EX-XX” and will be placed at the very end of your
current policy. We can move them after that where we prefer, usually before the rule that was under investigation

Debug the Machine Learning

tail -f /tmp/error_SecRulesLearn

© 2020 Palo Alto Networks, Inc. 43

Push Modified Security Policies back to PAN-OS Devices
Generate API Requests

After you reviewed the modified security rules and ready to push the changes back to the PAN-OS device, you will
go to “Export” ->”API Output Manager”, click on the “[Step1] Generate API Requests” to generate all the API
calls. By default the mode will be “Atomic” , here are the details of each mode:

● Mega: generates one single API call containing the complete XML configuration.
● Atomic: generate one API call for each section in the configuration, such as address objects section,
service object section, etc. In case of multiple DGs, we may obtain one API call for each section in the DG.
● Subatomic: generate an API call for every element in a configuration. For instance, one API call for each
address object in a configuration. This may be required when only certain objects should be submitted to
the device.
● Clear: generate API calls to delete the content of the desired section

© 2020 Palo Alto Networks, Inc. 44

Send API Requests
You can then select the API calls you would like to send to the PAN-OS device by selecting the checkbox in front
of the API calls , the order of the API calls are listed in the “Id” column , you will start with the API calls with lower
Id, for example, if the changes in Expedition include below new objects and policies, you will push those API calls
one by one per below order:

Tag
Address objects
Address-Group objects
Service objects
Service-Group objects
Security Rules

You can filter the section of the configuration by selecting the section on the filter dropdown menu. check on the
API calls you would like to send to the PAN-OS device, click on “[Step 2] Send API Requests”

Select the Device you want to send the API call to , then click on”SEND”

© 2020 Palo Alto Networks, Inc. 45

After the command , the Screen will refresh with the status of the API calls, in the case of successful API call, th
status will show “command succeeded “ like below screenshot:

After you send all the required API calls, you can then verify changes in the PAN-OS device and Commit the
changes.

© 2020 Palo Alto Networks, Inc. 46

Revision History
Date Revision Comment
Nov,5, 2020 A First release of the documentation
Apr, 11, 2022 B Adding Splunk app info

© 2020 Palo Alto Networks, Inc. 47