Document point-in-time nature of /_status/tenant_ranges endpoint results

[abarganier]

Is your feature request related to a problem? Please describe.
The /_status/tenant_ranges endpoint provides metadata about ranges for a tenant that exist at the time the request was processed.

The initial tenant creation process involves the creation of initial ranges, followed by an asynchronous process that splits ranges. If a user were to make a /_status/tenant_ranges immediately after tenant creation, it's possible that the client would receive a response containing a fewer number of ranges than they would receive if they made another request a short time after. This would be due to the asynchronous nature of this range splitting process.

Ranges are forever in a state of flux. Splitting, compaction, etc. are common operations that happen to ranges, and so this behavior should be intuitively expected by those who understand this. However, we should make this limitation clear in the documentation, to avoid potential confusion in scenarios such as the example one I gave above.

Describe the solution you'd like
Update documentation to note this characteristic of the /_status/tenant_ranges endpoint.

Additional context
Informs #97386

Jira issue: CRDB-25678

Informs: #92979

[blathers-crl]

Hi @abarganier, I've guessed the C-ategory of your issue and suitably labeled it. Please re-label if inaccurate.

While you're here, please consider adding an A- label to help keep our repository tidy.

_{🦉 Hoot! I am a Blathers, a bot for CockroachDB. My owner is dev-inf.}

[andf-crl]

Created DOC-7271 to track this.

[lnhsingh]

Removing this from end-user-facing known limitations since this is an internal interface.
The docs for the internal interface is generated from code in cockroachdb/cockroach and can be updated to include info there. https://github.com/cockroachdb/cockroach/blob/master/docs/generated/http/full.md#tenantranges