Horton

Re: BRIN indexes - TRAP: BadArgument

Amit Langote <amitlangote09@gmail.com>

View thread View on postgr.es
From: Amit Langote <amitlangote09@gmail.com>
To: Fujii Masao <masao.fujii@gmail.com>
Cc: Alvaro Herrera <alvherre@2ndquadrant.com>, Jeff Janes <jeff.janes@gmail.com>, Robert Haas <robertmhaas@gmail.com>, Heikki Linnakangas <hlinnakangas@vmware.com>, Erik Rijkers <er@xs4all.nl>, Emanuel Calvo <3manuek@esdebian.org>, Simon Riggs <simon@2ndquadrant.com>, Nicolas Barbier <nicolas.barbier@gmail.com>, Claudio Freire <klaussfreire@gmail.com>, Josh Berkus <josh@agliodbs.com>, Andres Freund <andres@2ndquadrant.com>, Pg Hackers <pgsql-hackers@postgresql.org>
Date: 2014-11-10T01:53:55Z
Lists: pgsql-hackers 
On Sun, Nov 9, 2014 at 10:30 PM, Fujii Masao <masao.fujii@gmail.com> wrote:
> On Sat, Nov 8, 2014 at 4:56 AM, Alvaro Herrera <alvherre@2ndquadrant.com> wrote:
>>
>> I just pushed this, after some more minor tweaks.
>
> Nice!
>
>> Thanks, and please do continue testing!
>
> I got the following PANIC error in the standby server when I set up
> the replication servers and ran "make installcheck". Note that I was
> repeating the manual CHECKPOINT every second while "installcheck"
> was running. Without the checkpoints, I could not reproduce the
> problem. I'm not sure if CHECKPOINT really triggers this problem, though.
> Anyway BRIN seems to have a problem around its WAL replay.
>
> 2014-11-09 22:19:42 JST sby1 WARNING:  page 547 of relation
> base/16384/30878 does not exist
> 2014-11-09 22:19:42 JST sby1 CONTEXT:  xlog redo BRIN/UPDATE: rel
> 1663/16384/30878 heapBlk 6 revmapBlk 1 pagesPerRange 1 old TID (3,2)
> TID (547,2)
> 2014-11-09 22:19:42 JST sby1 PANIC:  WAL contains references to invalid pages
> 2014-11-09 22:19:42 JST sby1 CONTEXT:  xlog redo BRIN/UPDATE: rel
> 1663/16384/30878 heapBlk 6 revmapBlk 1 pagesPerRange 1 old TID (3,2)
> TID (547,2)
> 2014-11-09 22:19:47 JST sby1 LOG:  startup process (PID 15230) was
> terminated by signal 6: Abort trap
> 2014-11-09 22:19:47 JST sby1 LOG:  terminating any other active server processes
>

I could reproduce this using the same steps. It's the same page 547
here too if that's any helpful.

Thanks,
Amit

Commits

  1. Refactor per-page logic common to all redo routines to a new function.

  2. Reduce use of heavyweight locking inside hash AM.

  3. Scan the buffer pool just once, not once per fork, during relation drop.

  4. Major patch from Thomas Lockhart <Thomas.G.Lockhart@jpl.nasa.gov>