Nikolay
Nikolay
The bugfix was included into [v1.102.24](https://docs.victoriametrics.com/changelog/#v110224), [v1.110.12](https://docs.victoriametrics.com/changelog/#v111012) and [v1.120.0](https://docs.victoriametrics.com/changelog/#v11200).
The feature was included into [v1.105.0](https://docs.victoriametrics.com/changelog/#v11050)
Most probably, issue was introduced at the commit 658cb2409db415a12a4ddc00c85f07be1a9099ed It incremented indexDB ref for usage, but `indexDB.MustClose()` method doesn't wait for refCount == 0. It may lead to the incorrect...
> > Most probably, issue was introduced at the commit 658cb2409db415a12a4ddc00c85f07be1a9099ed > > so this bug existed even before open-sourcing of VictoriaMetrics? Couldn't provided commit. https://github.com/VictoriaMetrics/VictoriaMetrics/blame/baaa88001e3dbd55b0b4359d0c51ebb78a9500a6/lib/storage/index_db.go#L276 For now, it's hard...
The bugfix was included into [v1.132.0](https://docs.victoriametrics.com/changelog/#v11320), [v1.122.11](https://docs.victoriametrics.com/changelog/#v112211) and [v1.110.26](https://docs.victoriametrics.com/changelog/#v111026).
The partial bugfix was included into v1.132.0 release, vmselect doesn't cache `rate` function results any more. The proper fix will be to return back caching for `rate` function later.
The feature was included into [v1.120.0](https://docs.victoriametrics.com/changelog/#v11200).
The bugfix was included into [v1.132.0](https://docs.victoriametrics.com/changelog/#v11320), [v1.122.11](https://docs.victoriametrics.com/changelog/#v112211) and [v1.110.26](https://docs.victoriametrics.com/changelog/#v111026).
Hello, I also recommend to check persistent disk metrics for your installation. Maybe it hit some gce limit https://cloud.google.com/compute/docs/disks/performance#performance_limits VictoriaMetrics exposes the following disk throttling metrics itself: ``` rate(vm_filestream_read_duration_seconds_total) rate(vm_filestream_write_duration_seconds_total)...
> Key findings: Write latency reached 0.03–0.06 → ~30–60 ms/sec. Read latency spiked to 0.5 → ~500 ms/sec. > > These spikes align with CPU saturation periods. > > It's...