Vladimir Smirnov

Also, protocol level support can be added without introducing backward incompatibility. If you add a new field for timestamps to current protocol and add a flag for the request type,...

You might also want to grab those changes alongside the original PR: https://github.com/go-graphite/carbonapi/commit/5f4972d48a31e392e8bad89dcdb496f472c4b379 https://github.com/go-graphite/carbonapi/commit/2406fe74ab136c58fe6ddd34f08787693a25bf6a

My point was that after the PR was accepted, I've done some changes to it, which you may or may-not want to port. Completely up to you to decide which...

So far I've for myself implemented simple debug menu which just prints out Status register statuses during initialization at various points (as dec10 values) and all status (status0, status1, status0a,...

M... I no longer remember what was the description of status registers in datasheet, but even raw dump + datasheet was already helpful for me to confirm that it was...

I'm planning to finally fix that as part of #742 That should refactor how keep-alive, etc. is used ( I think one of the problems was with current approach is...

Thanks. Yeah, as it seems not a lot of people need it, it doesn't need to be especially efficient if that would take too much time to implement.

Can you also grab and share memory profile?

> @Civil I don't use go-carbon on heavy load. doMultipleRequestsIfSplit: false is recomemended for this ? It is niche. To have better results you need to have: 1. multiple go-carbon...

I think 200 OK can still be returned as a result if one backend replies. Also https://github.com/go-graphite/carbonapi/blob/main/doc/configuration.md#notfoundstatuscode by default for not found status code is set to 200 (for historical...