Symptom

Alarm will be raised for rootfs filesystem at 80% after upgrading to XR 732 & higher version. LC/0/4/CPU0:2022 Aug 3 12:45:07.224 UTC: resmon[361]: %HA-HA_WD-3-DISK_ALARM_ALERT : A monitored device / ( rootfs:/ ) is above 80% utilization. Current utilization = 80. Please remove unwanted user files and configuration rollback points.

Conditions

tmpdir_cleanup.sh script is responsible for cleaning /tmp when it reaches 90% capacity (max is 63 Mb). However before this changed it stored archive on /misc/disk1 no matter if that partition is mounted (e.g. harddisk on RP) or not (LC) thus on LC it consumed rootfs space. This optimization will cause script to delete archived /tmp files on LC to preserve rootfs space.

Workaround

This issue can be temperarily cleared by removing below files (if any). 1. Disable attestation agent. This will stop creating those ima files. config-t)# attestation runtime-integrity cache job disable config-t)# commit 2. Remove cache and phu files. these are safe to remove. run ssh lc_xr <== Here "Y" is slot number. For LC 0/2/cpu0, Y=2 cd /misc/disk1/ima/ima_cache_node0-x-CPU0-xxxxxxxxx chattr -i ima-cache-x rm ima-cache-x 3. Delete phu.log files from below paths; cd /tmp ls -l phu.log rm phu.log cd /misc/disk1/tmp_archive/tmp ls -l phu.log rm phu.log

Further Problem Description