Hi. We are using the open source version of OpsCenter. We find it useful, but the disk space for OpsCenter metrics has been increasing and can sometime outgrow to 30-50G in a matter of a day or two. We do have a lot of keyspaces and column families.
Usually this dev cluster is quiet on the weekend except for some QA jobs or our weekend primary read-repair. Most importantly, I truncated all rollups early this morning and during a big compaction (with hundreds of pending tasks at one point), the metrics grew to ~13G. Can I say compaction activities can increase the metric disk usage growth significantly? I have seen this behavior quite often with compaction. Thanks. John
