[Enhancement]: Replica identity full on hypertables with enabled compression

[flexwende]

What type of enhancement is this?

API improvement

What subsystems and features will be improved?

Compression

What does the enhancement do?

Similar to the latest additions in release 2.12.0 (#6086) and the pull requests #5868 and #5515, make hypertables, that have compression enabled, also support replica identity full.

The current state (2.12.0, postgres 15) is:

ALTER TABLE test_table SET (timescaledb.compress=True);
ALTER TABLE test_table REPLICA IDENTITY FULL;

ERROR: operation not supported on hypertables that have compression enabled

Is that generally possible at all with the compressed data structure?
If so, is implementation planned?

Implementation challenges

No response

[JamesGuthrie]

@flexwende We've unblocked the REPLICA IDENTITY statement on hypertables with compression configured. This means that it should be possible to set the replica identity on a hypertable, which will propagate to the uncompressed chunks, but it will not affect the compressed hypertable and its chunks.

This should be part of the TimescaleDB 2.16 release. I don't know when it will be released, but probably in a few months.