Setup Juniper ISG NSRP cluster

This post describes how to rebuild a Juniper NSRP Cluster if the first Juniper firewall is already configured for NSRP.
Please make sure you have the following prerequisite on both Firewalls.

Minimum software and hardware requirements for configuring Active / Passive NSRP:

  • Firewall’s with identical ScreenOS versions and license keys
  • Firewall’s with identical hardware
  • At least one interface on each firewall to be configured in the HA zone, which will be used for carrying control channel information

Configuration steps on the unconfigured Firewall.

Configure the Interface(s) for HA

If possible it makes sense to use the same Interface as used on the other device.

set interface ethernet0/4 zone HA

Configure the NSRP cluster id:

set nsrp cluster id 1

Both firewalls in the cluster must have the same Cluster ID number.

IMPORTANT: Other NSRP firewall pairs on the same segment must have a different set of cluster ids. Once the cluster id is set to a value, all the security interfaces will become part of the VSD-group 0, by default.

Configure cluster name for NSRP:

To define a single name for all cluster members, type the following CLI command:

set nsrp cluster name <name_str>
set nsrp vsd-group id <number> priority <number>

IMPORTANT: Make sure that the desired STANDBY firewall has a HIGHER priority configured than the preferred master. The firewall with the lower priority will be the active master in the cluster!

Physical connection:

Connect only the HA link cable to the interface that is configured for the HA zone.

Check the nsrp cluster status:

You can check the nsrp cluster status via the configuration GUI or via CLI.

To check the NSRP cluster status via GUI connect to the actual master and navigate to Reports > System Log > Event and search for NSRP related entries.

To check the NSRP cluster status via CLI please connect to the actual standby via console. You can check the status with the following command:

get nsrp cluster

Synchronize the configurations:

To synchronize the configuration from the active to the standby unit, please connect to the standby unit via console and execute the following command:

exec nsrp sync global-config save

The following will be reported shortly after you enter the above command:

load peer system config to save
firewall-B(B)-> Save global configuration successfully.
firewall-B(B)-> Save local configuration successfully.
firewall-B(B)-> Done.
firewall-B(B)-> Please reset your box to let cluster configuration to take effect!

Reset the Firewall:

IMPORTANT:  If you are prompted to save the configuration after you enter the reset command, answer n (No).  Then, proceed with the reboot by answering y (Yes).

firewall-B(B)-> reset
firewall-B(B)-> Configuration modified.  Save? [y]/n n
firewall-B(B)-> System reset.  Are you sure? y/[n] y

Check if the configurations are in sync:

Please execute the following command via CLI (console connection) on the backup firewall to check if the configurations are in sync:

exec nsrp sync global-config check-sum

Physical connection:

Please connect all other interfaces in the correct order to the standby unit.

Initial manual failover:

exec nsrp vsd-group 0 mode ineligible

Command reference:

NSRP

get nsrp cluster Show cluster info
get nsrp monitor Show list of monitored interfaces
get nsrp vsd id 0 Show VSD id 0
get counters ha Show HA interface hardware counters
exec nsrp sync global-config check-sum Allows you to see if the cluster configs are syncronised
exec nsrp sync global save Sync’s the nodes.A reboot is required to complete the update.
exec nsrp vsd-group 0 mode Fails over the cluster. Run this command on the Master node.

Current Settings / Values

get envar get environment variable
get config get device configuration
get system get system information
get arp get arp cache
get route get routing table

Configuration re-sync Juniper ISG NSRP cluster

When the ISG cluster devices are correctly configured, they will start synchronizing the configuration. You can check if the configurations is in sync with the following command:

Cluster1:fw01(M)-> exec nsrp sync global-config check-sum     
configuration in sync

It can happen from time to time that the configuration of the cluster run out of sync:

Cluster1:fw01(M)-> exec  nsrp sync global-config check-sum 
Warning: configuration out of sync

To solve this issue, you should force the configuration sync (only on the backup device!!!!):

Cluster1:fw02(B)-> exec nsrp sync global-config save
load peer system config to save
Save global configuration successfully.
Save local configuration successfully.
done.
Please reset your box to let cluster configuration take effect!

Cluster1:fw02(B)-> reset 
System reset, are you sure? y/[n] y
In reset ...

After the reboot of the backup device, the cluster should be fully operational and in sync.

Note : when the device prompts you to save the config, enter “n” (no)

Strange VPN behaviour between Juniper ISG and ASA

Always something new what is not working, this time a site-to-site VPN tunnel between a Juniper ISG 2000 and a Cisco ASA 5520. After the initial set-up what a colleague did everything seemed to work fine with the VPN, the tunnel come up and communication was possible. What my colleague not did was to do some intensive testing and check in detail the configuration.

The problems began ones these VPN tunnels were used for production traffic, they worked fine for some hours or maybe days but then suddenly you were not able to send or receive any data over the tunnel and mostly it happen over night where it was difficult to troubleshoot. In the beginning everybody said that it was a problem within the internet but when it happened three or four times a week I got suspicious that its something wrong with the boxes or even configuration.

First thing was to check the configuration and quickly I found a difference in the IPSec Phase2 default timers, the Juniper uses 1 hour and Cisco has 8 hours. I was happy and thought I found the reason for the connectivity problems but then two days later it happened again… 🙁

There where some strange syslog messages about anti-replay checking failed:

ASA-4-402119 IPSEC: Received an ESP packet (SPI= 0xFBEFD496, sequence number= 0x1C24) from 80.210.78.10 (user=80.210.78.10) to 31.217.30.132 that failed anti-replay checking.

These messages always came when we had the connectivity problems and that could be only now the last reason. I have to say I didn’t really found much but there were some articles that Juniper doesn’t support Cisco Anti-Replay checking and that you need to disable that on the Cisco side.

Here a release note I found from Juniper but it’s about Cisco’s GET VPN:

http://www.juniper.net/techpubs/en_US/junos11.1/information-products/topic-collections/release-notes/11.1/index.html?topic-53319.html

In the end that was the only thing I found and I thought give it a try and disable the anti-replay checking on the ASA when we had the problems again. I did so I waited until the problem happened again and disable the anti-replay checking with the following command:

crypto ipsec security-association replay disable

It’s global command were you disable the checking for the whole device. Here some more information:

http://www.cisco.com/en/US/docs/security/asa/asa81/command/ref/c5.html#wp2200043

When I disabled the checking the communication was suddenly possible again and since then the problem never happened again so the anti-replay checking was the cause of the connection problems between the Juniper and the Cisco. Always something new to learn 🙂