James Guthrie

As noted in Promscale's [documentation](https://docs.timescale.com/promscale/latest/manage-data/delete-data/#deletion-by-series-with-http-api), the `delete_series` endpoint does not support deleting data with a time range: > NOTE > The start and end timestamp options are not currently supported....

@zongjiasushi it looks like you should be able to pass `true` as a second argument to `decompress_chunk` in order to tell it to skip uncompressed chunks (documentation [here](https://docs.timescale.com/api/latest/compression/decompress_chunk/#decompress-chunk)). I would...

I'm going to reproduce the analysis from the commit message here: The performance implications of this change are not entirely clear-cut. The following comparison between master and this change (benchmarks...

One theory for why this happens is worse escape analysis behaviour, causing more heap allocations. In order to test that theory, I created a standalone `main.go` which only uses the...

I decided to run the code in a "real-world" setup (ingesting ~200k samples/sec from ~5 million series over ~1000 metrics) to see if this difference was measurable. The following graph...

@brasno Thanks for the bug report. Can you give us an indication of the volume of metrics that you're trying to ingest (e.g. 100'000/s). Can you also see if there...

One important change to Promscale since I opened this issue is that the extension is now mandatory. I believe that there is still a use-case for the `install-extensions` flag: users...

I'd like to note that one of the main caveats with the change to `extension_sql_file` is that it requires setting the `name` attribute. This stems from the fact that we...

> I'm wondering if there is a way we can preserve the auto-naming if the user passes a `syn::LitStr` instead of an `syn::Expr`... I was wondering the same. I'm not...

@pankaj-83incs I ran into this recently as well. It seems as though the most recent chart versions fix the problem that you're running into, but those have not been released...