Help us improve your experience.

Let us know what you think.

Do you have time for a two-minute survey?

 
 

Software Upgrade in Multinode High Availability

Overview

SRX Series Firewalls deployed in an MNHA configuration can be upgraded with minimal disruption by sequentially upgrading each device. Depending on your device architecture, use one of the following CLI commands to initiate the Junos upgrade- request system software add or request vmhost software add .

From Junos OS Release To Junos OS Release Use Software Upgrade Method
20.4 Any release post 20.4 No
22.3 Next version of Junos OS Release Yes
  • Releases 22.4R1 and later are not compatible with earlier Junos OS releases for synchronizing sessions during a regular upgrade. Use the Isolated Nodes Upgrade Procedurein such cases.

  • Upgrading from 22.3 to the next release may cause brief traffic disruption.

  • You may see Peer Hardware Incompatible: SPU SLOT MISMATCH during upgrades from 21.4R1 onwards.

  • NAT sessions are not synced during interim upgrade stages in releases prior to 23.4R2.

  • Always upgrade both nodes to the same Junos OS version.

For information about upgrade and downgrade support for Junos OS releases, see Upgrade and Downgrade Support Policy for Junos OS Releases and Extended End-Of-Life Releases in Release Notes.

When you are upgrading SRX Series Firewalls in Multinode High Availability to Junos OS Release 22.4R1 or to a higher release, from an earlier Junos OS release, you can use the Isolated Nodes Upgrade Procedure. Junos OS Release 22.4R1 and higher releases are not compatible with earlier Junos OS releases for synchronizing sessions during a regular upgrade.

Before You Begin

Before performing an upgrade on an SRX Series device in MNHA) configuration, it is recommended to redirect the traffic away from the device in a controlled way. This can be done using one of the following methods:

  • Manual failover —Trigger a manual failover to shift traffic to the peer device.

  • Software upgrade mode —Temporarily configure the device with the following command:

    This command introduces a device failure with failure code SU (Software Upgrade). As a result, Services Redundancy Groups (SRG) 1 and above will transition to an Ineligible state (instead of Active or Backup) on the device being upgraded. This causes the associated traffic to automatically fail over to the other MNHA cluster member.

    Note: If your MNHA cluster is configured with only SRG0 and includes the install-on-failure-route option, you can still redirect traffic by using the set chassis high-availability software-upgrade configuration to move traffic off the device gracefully.

Software Upgrade

Preparation Checklist

Consider the following best practices when you plan your software upgrade:

  • Ensure both nodes are online and running the same Junos OS version. Check the current Junos OS software version on your device using the show version command.
  • Verify storage availability: show system storage
  • Check hardware status:
    • show chassis fpc pic-status
    • show chassis alarms
  • Ensure that there are no uncommitted changes.
  • Backup configuration and license keys.
  • Download the Junos OS image to /var/tmp on both devices.
  • Ensure your high availability setup is healthy, functional, and that the interchassis link (ICL) is up.

    show chassis high-availability information

  • Prepare your SRX Series Firewalls for an upgrade using the checklist available in .
Tip: We recommend that you perform software upgrades during a maintenance window.

For details on preparing your device for an upgrade, see Preparing for Software Installation and Upgrade (Junos OS).

Download Software

Download the Junos OS image from the Juniper Networks Support page on both SRX Series Firewalls and save it in the /var/tmp location. Example:

Upgrade Procedure

Follow the steps in this procedure to upgrade SRX Series devices configured in a Multinode High Availability (MNHA) setup. In this example, the cluster consists of two devices: srx-01 (currently active) and srx-02 (currently backup). The upgrade process begins with the backup node (srx-02), followed by the active node (srx-01), ensuring minimal service disruption.

  1. Ensure your Multinode High Availability setup is healthy, functional, and that the interchassis link (ICL) is up.

    On SRX-01 Device

    On SRX-02 Device

  2. Initiate the software upgrade process on the backup node (srx-02) and commit the configuration

    This command triggers a local failover for SRG0 and marks SRG1 (if present) as INELIGIBLE, allowing the peer node to take or retain the active role

  3. Verify the status of Multinode High Availability. The output shows Node Status: OFFLINE [ SU ], which indicates that the node is ready for the software upgrade. You can see that the status of the SRG1 has changed to INELIGIBLE.
  4. Confirm that the other device (srx-01) is in an active role and is functioning normally.

    The command output shows that the status of SRG1 is ACTIVE.

    Note that under the Peer Information section of the SRG1, the status is INELIGIBLE which indicates that the other node is in ineligible state.

  5. Install the Junos OS software on the SRX-02 device.
  6. Reboot the device using the request system reboot command after successful installation.
  7. Check the Junos OS version after reboot.

    The output confirms that the device is upgraded to the correct Junos OS version.

  8. Check status of the Multinode High Availability on the device.

    The output continues to display the node status as OFFLINE [ SU ] and SRG1 status as INELIGIBLE.

  9. Remove the software-upgrade statement and commit the configuration.

    When you remove the software-upgrade statement, the node failover state and any installed routes are cleared. Until this statement is removed, the node remains offline and all SRGs stay in the INELIGIBLE state. This effectively isolates the node from handling traffic during the upgrade, as long as the peer remains healthy.

  10. Check the Multinode High Availability status again to confirm that the device is online and the overall status is healthy and functioning.

    The output shows Node Status: ONLINE and SRG1 status as BACKUP, which indicates that the node is back online and is functioning normally in backup role.

  11. Check interfaces, routing protocols, routes advertised and so on to confirm that your setup is operating normally.

  12. Now you can proceed to upgrade the other device (SRX-01) using the same procedure.

Note:

(Optional) In case if you face any issues and cannot complete the upgrade, you can roll back the software on the device, and then reboot the system. Use the request system software rollback command to restore the previously installed software version.

Upgrade Software using install-on-failure-route

For setups using only SRG0 (without A/B state support), we recommend configuring the install-on-failure-route. This route can be referenced in route policies to advertise less preferred paths during software upgrade scenarios or node failures. In this method, you can divert the traffic by changing the route. Here, traffic can still go through the node and interface remains up.

  1. Create a dedicated custom virtual router for the route used for diverting traffic during the upgrade.

  2. Configure the install-on-failure-route statement for SRG0. Here, you have configured the route with IP address 10.39.1.3 as the route to install when the node fails.

    The routing table installs the route mentioned in the statement when the node fails.

  3. Configure a matching routing policy and define a policy condition based on the existence of routes. Here you include the route 10.39.1.3 as the route match condition for the if-route-exists.
  4. Create the policy statement to refer the condition as one of the matching term.

  5. Initiate software upgrade as mentioned in previous steps (Software Upgrade).

Deprecated Method (shutdown-on-failure interface)

Starting in Junos OS Release 24.3R1 onwards, the shutdown-on-failure functionality is deprecated— rather than immediately removed—to provide backward compatibility and an opportunity to bring your configuration into compliance with the new configuration. As a part of this change, the [set chassis high-availability services-redundancy-group 0 shutdown-on-failure interface-name] configuration statement deprecated.

Previously, traffic had to be diverted manually by shutting down interfaces. You can now use the software-upgrade command to keep the node offline and all SRGs in the INELIGIBLE state for the duration of the upgrade. This effectively isolates the node from handling traffic.

If you're using Junos OS 22.4 or earlier, we recommend using the legacy methods to divert traffic during the upgrade.