release-22.2: fix unstable sort in SSTSink

[adityamaru]

This is a forward port of #95446.

The focus of this patch is fixing the unstable sort described in #95445. Previously, we would sort only on the basis of the start key of two backup files returned as part of an ExportResponse. This could result in older revisions of a key getting sorted and flushed to the underlying SSTWriter before new revisions, resulting in out-of-order writes, causing the backup to fail. We saw this in the support issue cockroachlabs/support#1998 and have a test TestExportRevisionsWithTimestampPagination that failed under stress.

To fix this unstable sort we now use the following algorithm:

Sort on start key of the backup files.
If start key is the same, sort on end key of the backup files. If end key is the same, sort in descending order on the end key timestamp of the files as both files are guaranteed to only contain revisions of the same key.

Fixes: #95445

Release note (bug fix): Fixes a bug where a backup of keys with many revisions would fail with "pebble: keys must be added in order".

Release justification: high impact bug fix to prevent failing backups

[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 

[adityamaru]

friendly ping @stevendanna!