There are several useful statistics when trying to size data nodes for a bucket using the high level formula in the documentation:
The difficulty I see is that some of the statistics are available from the REST API, whilst others are only available via thecbstats CLI.
This poses an additional challenge, as you can use cluster-wide values from the REST API, whilst the CLI is only providing them on a per node basis.
Trying to mix and match these in a calculation is a bit messy.
In particular, it seems ep_value_size stat is only available via CLI.
Can anyone confirm this, maybe there is another REST end point I’m missing?
If not, is it a reasonable for me to open a feature request in JIRA to have the missing stat included in REST as well as CLI?