eddietanshuo

connection rate increase significantly after upgrade to 1.77.0

18

**Describe the bug** Hello @valyala, we find a new problem with the conn pool. To solve the issue https://github.com/VictoriaMetrics/VictoriaMetrics/issues/2508, we upgrade the vmselect to newest build. It seems the tcp...

A single vmselect pod receives different number of row samples and blocks each time from the vmstorage cluster with the same query expression and timerange.

2

**version** 1.69.0-cluster **Describe the bug** We have a grafana dashboard, it occasionally gets error response: `cannot select more than -search.maxSamplesPerQuery=1000000000` the dashboard is configured with following params: - query time...

assisted merges result in vmstorage slow query and ingest

1

**Describe the bug** Hello @valyala, another issue need your help. One of our vm cluster had searching and slow ingesting problem this noon. > slow ingesting  > slow searching...

`histogram_quantile` get incorrect result when samples miss.

10

**Describe the bug** Hello @valyala @YuriKravetc, we find a new issue in our product env. We get 2 datasource, the left one is our prometheus native storage and the right...

vmstorage: could we have a more flexible lsm parts limit.

3

**Is your feature request related to a problem? Please describe.**  We find the total lsm parts opened by vmstorage increased by a significant percentage on each 1st of a...

vmselect: should we limit the size of the connection pool to avoid livelock?

3

**Describe the bug** We meet the same issue with https://github.com/VictoriaMetrics/VictoriaMetrics/issues/2552 again, our number of tcp connections reached 30k (our hard limit of process is near 100k) when assisted merging happen....

https://github.com/gosnmp/gosnmp/blob/448fcd8a01ec358c87ca8aea78ba8b73ae123590/helper.go#L123C1-L124C1 Some device has zero size of OID value. Please let it be compatible with old devices.