Description of problem: Customer has recently FFUed from RHOSP13 to RHOSP16.1 and then enabled the legacy telemetry services and using collectd. after upgrade the collectd memory was growing rapidly and causing out of memory on the nodes. Some best practices were followed like removing the STF configurations and keping only legacy telemetry configs, removing extra plugins from collectd, configure WriteQueueLimitLow/WriteQueueLimitHigh, and increasing the time interval between metric collection. This has helped to improve the memory consumption problme significantly, but over the period of time it is observed that collectd memory usage is increasing in controller/compute/ceph nodes and ceilometer-agent memory usage is also increasing in controller node.
Closing this, since there was no response.
The needinfo request[s] on this closed bug have been removed as they have been unresolved for 500 days