Hello: We recently have increased the amount of log files processed by our managed cluster of 3 nodes.
We are now receiving messages of log trimming due to log event limits.
I have increased in the admin portal the maximum amount, but this is still being exceeded
My question is, how does Dynatrace distribute the load amongst the hosts in a managed cluster environment?
Does it always distribute it evenly? (If there are 4 servers, each server gets 1/4th of the logs to process?)
Or does it use an algorithm to determine the maximum amount based upon other factors?
I am asking this because the next node I want to add to the managed cluster I would prefer to taylor its capabilities towards whatever methods Dynatrace uses for log ingestion.
Solved! Go to Solution.
Maybe you can find some useful information here:
I hope it helps.
The load is distributed evenly among all cluster nodes. If you hit the limit you can either add a node or add more RAM to the cluster nodes. See details here:
Thank you Radoslaw_Szulgo.
This is great information. I have one follow up question.
If I increase the memory on the existing servers, or if I add a cluster member. Is a reboot required of the existing systems to re calcluate the maximum logs ingested? Or is this number recalculated on a regular basis or time interval? (In other words, how does Dynatrace Managed know to increase the max Log ingested after memory or a cluster node is added>
You can refresh the cluster level limit in the management console at any environment after the hw upgrade. This would be the cluster level log ingest capacity. It can be shared between the environments.
In this example there is a cluster with 3 hosts (66GB RAM and 8vcpu / host). Overall cluster level capacity is 26k/min and 20k allocated to this specific environment.
I hope it helps. Have a nice day.