Cluster sizing and hardware calibration

[jseldess]

Jesse Seldess commented:

This will be an internal product marketing engineering project first to create a sustainable well to generate good numbers and guidance. Eventually, this might make it into customer-facing docs.

See: https://docs.google.com/document/d/1RUdDcE6jQfA9QzMLFZ2ta2JWK1D-mJ7afABAluQXtAo/edit#heading=h.4chbev4vd0oz

Jira Issue: DOC-228

[jseldess]

@robert-s-lee, @timveil, @rkruze, @drewdeally, please give a thumbs up if the above makes sense; otherwise, add comments to help me clarify this work.

[jseldess]

In describing the characteristics of a workload, we should be sure to include expected txn retry ratios.

[jseldess]

Meeting schedule to discuss next steps here.

[jseldess]

@Amruta-Ranade, how much does this overlap, or not, with your performance scaling project? #5155

[Amruta-Ranade]

Significant overlap, I'd say.

[exalate-issue-sync]

Michael Wang (mwang1026) commented:
We should definitely do this. I think as we do more performance tests we should keep an eye on best practices here.