This article describes in details behind the reason for getting multiple alarm on rpd-sa-agent and other applications after the upgrade.

After performing the upgrade customer may get below alarms on the on the box.

root@juniper> show system alarms no-forwarding

4 alarms currently active

Alarm time        Class Description

2023-11-06 13:19:27 UTC Major Application commit-syncd fail on node Re0

2023-11-06 13:09:06 UTC Major Application dot1xd fail on node Re0

2023-11-06 13:09:06 UTC Major Application rpd-sa-agent fail on node Re0

2023-11-06 13:09:06 UTC Major Application rpdtmd fail on node Re0

Issue is due to invocation of "HwdPfePowerConfigValidation.validatePfePowerConfig" function as commit check with power on/off config. During run time when we apply pfe power on/off config, FPCx nodes are present in sysfs and we dont see any commit issue.

But during image upgrade, when system is coming up with new image config validation is invoked well before fpc sysfs nodes are created.
Due to which HwdPfePowerConfigValidation.py stuck in rsh and we got into issue.

In normal reboot, there is no commit validation hence we dont see issue in subsequent reboot.

Restoration Steps:
Please restart the corresponding daemon (for example, root@user>restart dot1xd) or reboot the node to restore services.

Workarround:
Please remove the 'set chassis fpc <fpc-slot> pfe <pfe-id> power (off|on)' configuration before upgrading the image and then configure it back post the upgrade is complete.

Issue is fixed via below PR:

https://prsearch.juniper.net/problemreport/PR1705725