Ganesh Vernekar

Somehow I missed this issue. Yes the snapshot does not mean it won't touch the WAL. There could be many reasons why new WAL files came in later without a...

https://github.com/prometheus/prometheus/pull/11340 for the docs

> > Simplify loadWAL and loadWBL in head_wal.go as much as possible (not directly related to the OOO work, since this structure of code existed before) > > As part...

I think we have a bug in a metric https://github.com/prometheus/prometheus/blob/73d805cf6c67cadf8e56d6fb8958d1408be869dc/tsdb/head_append.go#L317-L319 The histogram buckets assume seconds, but here most of the timestamps are going to be milliseconds. So we should be...

Catching up with emails now :) looks like I missed some discussions > @codesome Given we have vertical compaction in TSDB and we can have overlapping blocks, what would be...

I would as much try to avoid adding samples older than Head mint and bring vertical compaction into play in the upstream Prometheus, because (1) Code is already complex enough...

> Can you share those ideas Ganesh? It was one of my TODO for today :) I will share with you all once I have written it down.

This is an idea from the top of my head, needs more thought **The root problem** These 2 checks, [this](https://github.com/prometheus/prometheus/blob/6ff4814a492ae6f9e3f5cf7b022a823b631b7a95/tsdb/head.go#L1084-L1087) and [this](https://github.com/prometheus/prometheus/blob/6ff4814a492ae6f9e3f5cf7b022a823b631b7a95/tsdb/head.go#L1114-L1117) **Solution** Enable vertical compaction and querier and make...

> Can we elaborate? Here the "check" is checking of min valid time for the samples - where we discard samples before 1h. When I say make that check optional,...

Starting another TSDB on the side for non-backfilling purposes is a lot of complexity for the normal ingestion path. So I suggest we have a separate API itself for backfilling...