Cluster Perfomance History Keeps Filling Up

Question

When running enable-clusters2d we get a new volume called ClusterPerformanceHistory. This stores the cluster performance data that rhs.exe adds, so we can see cluster performance over time. The issue I am seeing is that some clusters we have to increase the size of the ClusterPerformanceHistory volume to as much as 250GB's and its still running out of space. Is there any guidance on how large to make ClusterPerformanceHistory or how to minimize the data being written to avoid running out of space? When ClusterPerformanceHistory runs out of space we no longer can query or use that data. Any help would be much appreciated.

Answer 1

This should not be filling up so quickly. Take a look at the Timeframes section in this doc, it will reduce the data set for data as it ages:
Performance history for Storage Spaces Direct | Microsoft Learn

You could use the -TimeFrame parameter to adjust the history. Or you may want to Stop-ClusterPerformanceHistory -DeleteHistory (then Enable again) to purge, in the event that there is aged data that for some reason is not being reduced.

You could also increase the size of the disk used to store the performance history, see this doc:
S2D ASHCI: Extend the Cluster Performance History Cluster Disk – MPECS Inc. Blog

If you do not care about the data being collected, just reduce the frequency. If you think the diminishing granularity is not working properly and want to troubleshoot further, your best bet will be to speak with support.