High CPU utilization by PMM agent

Percona Monitoring and Management (PMM) PMM 2.x
1

Description: High CPU utilization by PMM agent
We observed high CPU utilization by PMM agent in our MongoDB secondary servers.
After restarting the agent, it was resolved. In the pmm-agent logs we observed below sample logs. Similar logs were observed for multiple namespace and collections. After restart of pmm-agent, these errors are not observed

level=error msg="cannot get $collstats cursor for collection rest.tmp.mrs.logDetails_1739062843_1120: (Location40280) Unable to retrieve storageStats in $collStats stage: Collection [rest.tmp.mrs.logDetails_1739062843_1120] not found." collector=collstats" agentID=/agent_id/c16fe8c3-f086-4501-89a6-3fa40aa893fb component=agent-process type=mongodb_exporter

Version:
PMM 2.44
Mongo version: 4.0.28

2

Hello @ikon

It’s better to report this as a bug, As it needs more details to confirm the issue.

You can report it here with full details on the issue: https://perconadev.atlassian.net