release-21.2: sql: consistent aggregated timestamp when flushing sql stats

[Azhng]

Backport 1/1 commits from #71731.

/cc @cockroachdb/release

---

Previously, when SQL Stats are flushed to system table, the
aggregatedTs column for SQL Stats is calculated for each stats
entry individually. This means that if a flush starts near
the end of each hour, it is possible that different stats
rows will be assigned two different aggregated timestamp.
Currently, flusher flushes statement statistics first, and only
when statement statistics are flushed, transaction statistics
will be flushed into system table. This means that it is likely
the transaction statistics will get assigned a different
aggregatedTs than the statement statistics.
Consequentially, when the frontend fetches SQL Stats through
the CombinedStmtStats handler, the frontend default performs
range scan at 1 hour interval. This triggers a range scan
on the system table for that 1 hour range. This causes the
statement statisitcs that got assigned a different aggregatedTs
to be omitted from the result.

This commit changed the flusher to only compute aggregatedTs once
before the flush actually happen, and assign that aggregatedTs
too all stmt/txn stats rows. Statements executed in the same
aggregation interval can be looked up by the corresponding
statement fingerprint ID stored in transaction stats metadata.

Follow up to #71596

Release note: None

[blathers-crl]

Thanks for opening a backport.

Please check the backport criteria before merging:

• Patches should only be created for serious issues or test-only changes.
• Patches should not break backwards-compatibility.
• Patches should change as little code as possible.
• Patches should not change on-disk formats or node communication protocols.
• Patches should not add new functionality.
• Patches must not add, edit, or otherwise modify cluster versions; or add version gates.

If some of the basic criteria cannot be satisfied, ensure that the exceptional criteria are satisfied within.

• There is a high priority need for the functionality that cannot wait until the next release and is difficult to address in another way.
• The new functionality is additive-only and only runs for clusters which have specifically “opted in” to it (e.g. by a cluster setting).
• New code is protected by a conditional check that is trivial to verify and ensures that it only runs for opt-in clusters.
• The PM and TL on the team that owns the changed code have signed off that the change obeys the above rules.

Add a brief release justification to the body of your PR to justify this backport.

Some other things to consider:

• What did we do to ensure that a user that doesn’t know & care about this backport, has no idea that it happened?
• Will this work in a cluster of mixed patch versions? Did we test that?
• If a user upgrades a patch version, uses this feature, and then downgrades, what happens?

[cockroach-teamcity]

This change is 

[Azhng]

Closing in favor of #72941