Re: remove spurious CREATE INDEX CONCURRENTLY wait - Mailing list pgsql-hackers

From Tom Lane
Subject Re: remove spurious CREATE INDEX CONCURRENTLY wait
Date
Msg-id 3397538.1597158895@sss.pgh.pa.us
Whole thread Raw
In response to Re: remove spurious CREATE INDEX CONCURRENTLY wait  (James Coleman <jtc331@gmail.com>)
Responses Re: remove spurious CREATE INDEX CONCURRENTLY wait
List pgsql-hackers
James Coleman <jtc331@gmail.com> writes:
> Why is a CIC in active index-building something we need to wait for?
> Wouldn't it fall under a similar kind of logic to the other snapshot
> types we can explicitly ignore? CIC can't be run in a manual
> transaction, so the snapshot it holds won't be used to perform
> arbitrary operations (i.e., the reason why a manual ANALYZE can't be
> ignored).

Expression indexes that call user-defined functions seem like a
pretty serious risk factor for that argument.  Those are exactly
the same expressions that ANALYZE will evaluate, as a result of
which we judge it unsafe to ignore.  Why would CIC be different?

            regards, tom lane



pgsql-hackers by date:

Previous
From: Robert Haas
Date:
Subject: Re: Switch to multi-inserts for pg_depend
Next
From: Tom Lane
Date:
Subject: Re: Can I test Extended Query in core test framework