Re: Possible causes for database corruption and solutions

On 16/12/2009 3:54 PM, Florian Weimer wrote:
> * Michael Clark:
>
>> The solution to the problem seemed to be to change the value for the
>> wal_sync_method setting to fsync_writethrough from the default of fsync.
>> I was curious if there were perhaps any other reasons that we should look
>> at? Or if there may be other alternatives to changing the wal_sync_method
>> setting.
>
> Fsync and related settings only matter if the operating system (not
> just the database) crashes. Does this happen frequently for you?

When you're dealing with end users who have machines running
god-knows-what kinds of awful hardware drivers and with no power
protection, then I expect it does. Add laptop users with ageing/flakey
batteries, laptops let go flat after they go into powersave suspend,
etc, and you're sure to see plenty of cases of apparent crashes.

--
Craig Ringer