GitHub user PragmaTwice edited a comment on the discussion: TimeSeries Proposal
Also, to improve performance, I suggest: - `size` fields contains chunk count, `chunkCount` can be removed. - `firstTimestamp` `lastTimestamp` can be removed. So that we don't need to rmw metadata for each `TS.ADD`. (maybe more than 50% speedup for `TS.ADD`.) If users query `TS.INFO`, we can return an approximate size = `chunk count * M`. (And we can provide a new command for query the precise sample size.) GitHub link: https://github.com/apache/kvrocks/discussions/3044#discussioncomment-13730081 ---- This is an automatically sent email for [email protected]. To unsubscribe, please send an email to: [email protected]
