High IO Wait on PostgreSQL Master - Possible PMM2 Issue?

Percona Monitoring and Management (PMM)
1

Hi,

We have a PostgreSQL cluster in a master/slave architecture. We deployed PMM2 in a container and registered the PMM agents without any issues.

However, after two months of great usage, we encountered two incidents where we suspect PMM might be the cause. In both cases, we observed unusually high I/O wait on the PostgreSQL master server. Unfortunately, I wasn’t able to capture the exact process responsible for this.

Has there been any known issue in the past related to PMM2 causing high I/O wait?

Thanks.

2

Hello,

what version of PMM Client is used?
How many DBs exist there?
What’s configured metrics resolution?
Do you have any custom query added?