Re: "Leaking" disk space on FreeBSD servers

From: Achilleas Mantzios <achill(at)matrix(dot)gatewaynet(dot)com>
To: Vick Khera <vivek(at)khera(dot)org>
Cc: pgsql-general <pgsql-general(at)postgresql(dot)org>
Subject: Re: "Leaking" disk space on FreeBSD servers
Date: 2013-03-20 14:47:48
Message-ID: 2123801.14EdYEJBbr@smadev.internal.net
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

Of course, but does it make sense for you to pay the ~ 5%/day performance penalty for the ~0.5%/year chance of having your system crush?
Unless your FreeBSD server is stuffed with exotic gamer hardware, i don't see the likehood of crush getting larger than that.

On Τετ 20 Μαρ 2013 10:39:58 Vick Khera wrote:

On Wed, Mar 20, 2013 at 10:34 AM, Achilleas Mantzios <achill(at)matrix(dot)gatewaynet(dot)com> wrote:

regarding journaling, there is the counter argument that you do not need to do the same job twice,
in the sense that we already spend a considerable amount of time retaining the WAL in postgresql,
no need to redo the same on FS level.

There's a difference in the file system integrity and the DB integrity. PG will keep the DB integrity just fine without the file system journaling. The journaling just makes recovery from crash that much faster. ie, running fsck on 6TB of disk storage takes a LONG time, sometimes hours, but with the journal enabled, it takes a few seconds.

-
Achilleas Mantzios
IT DEV
IT DEPT
Dynacom Tankers Mgmt

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Dan Thomas 2013-03-20 15:15:23 Re: "Leaking" disk space on FreeBSD servers
Previous Message jg 2013-03-20 14:47:34 Re: File Fragmentation