Roman Khavronenko

@valyala @stevenbeaullieu please see https://github.com/VictoriaMetrics/VictoriaMetrics/pull/3181

Hi @yzhengwei! As @laixintao said, these are two distinct time series for VictoriaMetrics. It just stores what you put in it. If you think you don't want some of these...

> Why does Vmagent automatically find such similar metrics ? vmagent is a metrics collector. It is not up to vmagent to decide which metrics to scrape or not. It...

Hello! Both metrics are produced by vmalert to remain complaint with Prometheus ecosystem. It is assumed that people already know about Prometheus and how it does the alerting thing. This...

Thanks for detailed report. Memory usage by vmselect looks like a memory leak. I assume some of the caches grows constantly, which makes lookups slower and slower. We need to...

Thanks for pprof. I don't see anything suspicious in it. How long it worked since the last restart? How much more memory it started to use since the the last...

> Btw do you know 2 groups is faster than one group? That shouldn't be a case. With `concurrency > number of rules`, the evaluation duration should be equal to...

> Hi this is the request per minutes Thanks! Looks good. I was thinking maybe vmalert for some reason increases requests rate because of slow responses. But looks like everything...

@laixintao is it possible to collect cpu and mem profiles right after the next vmselect restart, and then collect them once more when vmselect will start to get slower?

The `maxSearchQuerySize` is defined to 1MB, it means your query exceeds that size... Are there any other options to reduce the query size? Such a big regex expression as in...