| From: | Tomas Vondra <tomas(at)vondra(dot)me> |
|---|---|
| To: | Ramakrishna Reddy Nandyala <rkreddy(dot)n(at)gmail(dot)com>, pgsql-hackers(at)lists(dot)postgresql(dot)org |
| Subject: | Re: Page verification failed, calculated Checksum 1065 but expected 42487 |
| Date: | 2026-01-06 14:59:12 |
| Message-ID: | b3a31220-2946-47b1-a020-67bb0808d684@vondra.me |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On 1/6/26 14:59, Ramakrishna Reddy Nandyala wrote:
>
> Dear Team,
>
> We are using pg12 with rhel8 and Cybertrust Transperent encryption (CTE)
> as encysution,
I'm not sure what "Cybertrust Transparent Encryption" is exactly. I've
only found "CipherTrust Transparent Encryption" from Thales, but I have
no idea if that's what you're using ...
> recently we had faced a peculiar issue where all of sudden
> backup is getting failed with checksum validation failure for few
> objects. We have to truncate/restore the data from the backup in order
> to resume the backup and able to access the objects,
> we would like to know is there any known issues with respect to this
> combination of infra, any leads would be appreciated to understand the
> root cause of the issue.
>
I'm afraid that's impossible to answer without much more information. So
many things in various layers could have caused this. It could be a
Postgres bug, or maybe it's something in the transparent encryption
system, or maybe it's a storage issue. Who knows?
I suggest you talk to authors of the encryption solution, and try to
figure it out with them. I suppose you're paying them for the solution
and support, and they are probably in the best position to help you.
But they'll need some information about the problem too, I guess. You
haven't even told us which Postgres version you're running, on what OS,
how much data is there, and so on.
What I usually do in data corruption cases is to look for "weird" things
since the last successful operation (before hitting the issue). It might
be a crash/recovery of the database, storage issues, unexpected reboot,
or any other thing that does not happen often. It's hard to re
regards
--
Tomas Vondra
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Heikki Linnakangas | 2026-01-06 15:01:25 | Re: A few patches to clarify snapshot management, part 2 |
| Previous Message | Matthias van de Meent | 2026-01-06 14:58:05 | Re: Refactor PROCLOCK hash table into partitioned list allocator |