Re: Possible causes for database corruption and solutions - Mailing list pgsql-general

From Michael Clark
Subject Re: Possible causes for database corruption and solutions
Date
Msg-id bf5d83510912161108q1204be5br35da5bd940faefca@mail.gmail.com
Whole thread Raw
In response to Re: Possible causes for database corruption and solutions  (Greg Smith <greg@2ndquadrant.com>)
List pgsql-general


On Wed, Dec 16, 2009 at 2:05 PM, Greg Smith <greg@2ndquadrant.com> wrote:
Michael Clark wrote:
On Wed, Dec 16, 2009 at 11:25 AM, Greg Smith <greg@2ndquadrant.com> wrote:
Florian Weimer wrote:
I hope that Mac OS X turns off write caches on low battery.
 

I've never heard of such a thing.  The best you can do is try to push the system into hibernation instead of going down hard.  That *should* clear any disk caches as part of the graceful shutdown.  But you're relying on a relatively fragile system now, once the battery is quite low who knows if that will even execute in the window of time you have left.


And at this point it is not unreasonable to expect the user to perform some sort of action. Shutdown before the power completely dies, or plug in the power.

Sure, but the point is that if your database will get corrupted if that doesn't happen, that's a mistake because sooner or later every such portable system is going to end up there and not get plugged in on time.

Totally agree, not having PG setup properly and expecting the user to not fall into this problem is irresponsible.

I was speaking more generally on the OS level.  How aggressive should we expect an OS to be in a case like this where some responsibility should lie with the user. 


pgsql-general by date:

Previous
From: Greg Smith
Date:
Subject: Re: Possible causes for database corruption and solutions
Next
From: Erwin Brandstetter
Date:
Subject: How to get text for a plpgsql variable from a file.