On Mon, Jun 11, 2018, 10:22 PM Rui DeSousa <rui.desousa@xxxxxxxxxx> wrote:
> On Jun 11, 2018, at 9:28 AM, pavan95 <pavan.postgresdba@xxxxxxxxx> wrote:
>
> I'm searching a way to detect postgresql corruption on a daily basis. Please
> provide me
>
If you haven’t already (show data_checksums); I would recommend turning on data checksums to allow Postgres to detect i/o data corruption issues.
If it’s not enabled then you’ll have to reinit your database: initdb —data-checksums …
I would be very concerned why you feel the need to check for corruption and would question your platform. The only time I’ve seen corruption was due to a disk subsystem problem which data-checksums should help flush out for you or a Postgres bug which would normally gets bubbled up via the application and/or the logs.
Make sure you have good backups; there really is no script that going to solve a faulty disk subsystem or other disaster.
Thanks Rui,
Actually found some issues like segmentation fault with sigssv 11. I'm concerned about to make my database healthy against critical problems mainly in case of production environment. Also what's the instant thing that one needs to do in case of corrupted data.
Also it will be greatful if you can suggest me a script to find fragmentation
Regards,
Pavan
