This article explains a workaround for upgrading SRX CPE devices via CSO when upgrading from a specific platform version.

If upgrade is done between the following versions, an issue maybe occur due to PR1537386- AMR packets not going through expected path after interface flap.

• 19.3R2-S5 --> 20.2R3-S2
• 20.2R2-S3 --> 20.2R3-S2
• 19.3R2-S4 --> 20.4R2-S2
• 20.2R2-S3 --> 20.4R2-S2

The APBR issue as described in PR1537386 causes core while ISSU process is underway. The core causes the priorities to be 0, leading to ISSU failure because it needs priority to clear for the secondary to take over traffic and primary reboot to upgrade.

Workaround:

1. Stage Image on Device by navigating to Home > Resources > Images. Select the image and click on "Stage” option (image gets downloaded to primary node).

2.  Access Primary node by navigating to Home > Resources >Devices. Select the device radio button and click "More". Then select "Remote Console” to connect to the device.

3. Check the image copied to /var/tmp/ and copy the image to Secondary node: file copy /var/tmp/ <file name> node<0/1>:/var/tmp/

4. Login to Secondary node using:

   request routing-engine login node <0/1>

5. Upgrade Secondary Node: Execute the following command manually on secondary node:

   request system software add <image path> no-validate reboot

6. After rebooting Secondary node and it has is started, exit back to Primary node.

7. Upgrade Primary node: Execute the following command manually on primary node

   request system software add <image path> no-validate reboot

   Note: This is the stage where traffic disruption may occur since both nodes are rebooting to be upgraded.

8. Once the cluster is built up, login to POSTMAN and install telemetry agent using following API:

    https://<CSO IP/Hostname>/fmpm-provider/device_agent_install
     
   {
   "input": {
   "device_uuid":<device_uuid>
   }

1. Navigate to Home > Resources > Device
2. Gather the cluster UUID and punch it at " <device_uuid> " in the API request in step 8