Slow PITR restore

From: "Joshua D(dot) Drake" <jd(at)commandprompt(dot)com>
To: pgsql-general(at)postgresql(dot)org
Subject: Slow PITR restore
Date: 2007-12-12 02:18:32
Message-ID: [email protected]
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general pgsql-hackers

Hello,

I have an interesting thing going on here...

2007-12-11 21:10:36 EST LOG: restored log file
"00000001000007CC00000012" from archive
2007-12-11 21:10:47 EST LOG: restored log file
"00000001000007CC00000013" from archive
2007-12-11 21:11:02 EST LOG: restored log file
"00000001000007CC00000014" from archive
2007-12-11 21:11:14 EST LOG: restored log file
"00000001000007CC00000015" from archive
2007-12-11 21:11:26 EST LOG: restored log file
"00000001000007CC00000016" from archive
2007-12-11 21:11:39 EST LOG: restored log file
"00000001000007CC00000017" from archive
2007-12-11 21:11:54 EST LOG: restored log file
"00000001000007CC00000018" from archive
2007-12-11 21:12:13 EST LOG: restored log file
"00000001000007CC00000019" from archive
2007-12-11 21:12:27 EST LOG: restored log file
"00000001000007CC0000001A" from archive
2007-12-11 21:12:42 EST LOG: restored log file
"00000001000007CC0000001B" from archive
2007-12-11 21:12:54 EST LOG: restored log file
"00000001000007CC0000001C" from archive
2007-12-11 21:13:08 EST LOG: restored log file
"00000001000007CC0000001D" from archive
2007-12-11 21:13:21 EST LOG: restored log file
"00000001000007CC0000001E" from archive
2007-12-11 21:13:36 EST LOG: restored log file
"00000001000007CC0000001F" from archive
2007-12-11 21:13:45 EST LOG: restored log file
"00000001000007CC00000020" from archive
2007-12-11 21:13:52 EST LOG: restored log file
"00000001000007CC00000021" from archive
2007-12-11 21:14:01 EST LOG: restored log file
"00000001000007CC00000022" from archive

Now I understand that restoring log files can be slow but this is a big
machine.

32Gig of Ram
28 Disks (two volumes each raid 10)
Using two different table spaces (over each volume)
8 Opterons.

The machine has a production configuration that can doing several
hundred thousand transactions an hour without so much as having to
stretch first. It seems to me that this machine should be *cranking*
through these transactions. Am I just being my usual uptight, impatient
self?

Sincerely,

Joshua D. Drake

Responses

Browse pgsql-general by date

  From Date Subject
Next Message D. Dante Lorenso 2007-12-12 02:23:20 Function to convert from TEXT to BYTEA?
Previous Message Matthew Dennis 2007-12-12 02:01:09 Re: Creating Aggregate functions in PLpgSQL

Browse pgsql-hackers by date

  From Date Subject
Next Message Alvaro Herrera 2007-12-12 02:39:26 [[email protected]: Re: [pgsql-es-ayuda] SLL error 100% cpu]
Previous Message Tom Lane 2007-12-12 00:59:10 Re: archive_command failures report confusing exit status