Horton

Re: pg_dump: sortDumpableObjectsByTypeName() doesn't always do that

Andrew Dunstan <andrew.dunstan@2ndquadrant.com>

View thread View on postgr.es
From: Andrew Dunstan <andrew.dunstan@2ndquadrant.com>
To: Tom Lane <tgl@sss.pgh.pa.us>, Jacob Champion <pchampion@pivotal.io>
Cc: pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2018-08-06T19:16:56Z
Lists: pgsql-hackers 

On 08/06/2018 03:13 PM, Tom Lane wrote:
> Jacob Champion <pchampion@pivotal.io> writes:
>> We recently ran into an issue in pg_dump that caused the initial
>> sort-by-name pass to return incorrect results. It doesn't seem to
>> affect overall correctness, since the later toposort pass takes care
>> of dependencies, but it does occasionally cause a spurious diff in
>> dump output before and after a pg_upgrade run.
> Do you mean "incorrect results", or just "unstable results"?
> If the former, what's incorrect about it?
>
> 			


I'd also like to see a test case.

We should perhaps have a more representative set of data for the upgrade 
testing, both same version and cross-version, which judge success by 
comparing pre and post dumps, but I don't recall ever seeing this.

cheers

andrew

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

Commits

  1. Ensure pg_dump_sort.c sorts null vs non-null namespace consistently.

  2. Last-minute updates for release notes.