Horton

Re: [DOC] Document concurrent index builds waiting on each other

Alvaro Herrera <alvherre@2ndquadrant.com>

View thread View on postgr.es
From: Alvaro Herrera <alvherre@2ndquadrant.com>
To: Andres Freund <andres@anarazel.de>
Cc: James Coleman <jtc331@gmail.com>, pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2020-08-05T02:11:09Z
Lists: pgsql-hackers

Attachments

On 2020-Mar-25, Andres Freund wrote:

> What I was thinking of was a new flag, with a distinct value from
> PROC_IN_VACUUM. It'd currently just be specified in the
> GetCurrentVirtualXIDs() calls in WaitForOlderSnapshots(). That'd avoid
> needing to wait for other CICs on different relations. Since CIC is not
> permitted on system tables, and CIC doesn't do DML on normal tables, it
> seems fairly obviously correct to exclude other CICs.

Hmm, that does work, and seems a pretty small patch -- attached.  Of
course, some more commentary is necessary, but the theory of operation
is as Andres says.  (It does not solve the vacuuming problem I was
describing in the other thread, only the spurious waiting that James is
complaining about in this thread.)

I'm going to try and poke holes on this now ... (Expression indexes with
falsely immutable functions?)

-- 
Álvaro Herrera                https://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services

Commits

  1. Call out vacuum considerations in create index docs

  2. Document concurrent indexes waiting on each other

  3. snapshot scalability: Don't compute global horizons while building snapshots.

  4. Remove PROC_IN_ANALYZE and derived flags

  5. Improve performance of get_actual_variable_range with recently-dead tuples.