How to update benchmark results? (for Hyper)

[vogelsgesang]

We are about to release a new version of Hyper (probably later today) and would like to also update the ClickBench results.

I noticed that #993 put new automation in place.
Would you like us to still measure ourselves and create pull requests containing the new results?
Or should updates use the new benchmark-manual.yml Github flow? (We can't trigger flows ourselves, though, so we would need your support doing so)
Or should we pin the Hyper version in https://github.com/ClickHouse/ClickBench/blob/main/hyper/install, such that version bumps are done via explicit pull requests, which will trigger the benchmark-pr.yml Github flow?

[StefanSteiner]

The new version of the Hyper API for August 2026 is now available through: pip install tableauhyperapi

[rschu1ze]

Would you like us to still measure ourselves and create pull requests containing the new results?

We now run all new and updated submissions unconditionally through the automation before merging them. You don't need to mess with the tags or the mechanics of the automation - we are happy to do that for you.

Your PR may include measurements (result xml files) or not, it's optional. Including results can be useful to ensure on your side that the scripts work.

Or should we pin the Hyper version in https://github.com/ClickHouse/ClickBench/blob/main/hyper/install, such that version bumps are done via explicit pull requests, which will trigger the benchmark-pr.yml Github flow?

I'd say, if you bump the version in the install script and send that as a PR, it will be just fine. Thanks.

[vogelsgesang]

#1308 - let's see if that works smoothly with the new Github flows 🙂