Postgres Pro
Downloads
Home   >   mailing lists

Re: [PATCH] Use indexes on the subscriber when REPLICA IDENTITY is full on the publisher - Mailing list pgsql-hackers

From Önder Kalacı
Subject Re: [PATCH] Use indexes on the subscriber when REPLICA IDENTITY is full on the publisher
Date
Msg-id CACawEhUhn4bMuVeBQ30FM-OEtfy7+_UvTMxrYqcCASXEq08C4g@mail.gmail.com
Whole thread Raw
In response to Re: [PATCH] Use indexes on the subscriber when REPLICA IDENTITY is full on the publisher  (Andres Freund <andres@anarazel.de>)
List pgsql-hackers
Tree view
Hi Andres, Amit, all 

I think the case in which the patch regresses performance in is irrelevant in
practice. 

This is similar to what I think in this context.

I appreciate the effort from Shi Yu, so that we have a clear understanding on the overhead.
But the tests we do on [1] where we observe the regression are largely synthetic test cases
that aim to spot the overhead. 

And having an index over thousands
of non-differing values will generally perform badly, not just in this
context.

Similarly, maybe there are some eccentric use patterns that might follow this. But I also suspect
even if there are such patterns, could they really be performance sensitive?


Thanks,
Onder KALACI

 

pgsql-hackers by date:

Previous
From: Aleksander Alekseev
Date:
Subject: Re: RFC: logical publication via inheritance root?
Next
From: Önder Kalacı
Date:
Subject: Re: [PATCH] Use indexes on the subscriber when REPLICA IDENTITY is full on the publisher