Re: basebackup checksum verification

Hi,

On 2019-03-26 20:18:31 -0400, Stephen Frost wrote:
> > >>I thought Robert's response was generally good, pointing out that
> > >>we're talking about this being an issue if the corruption happens in a
> > >>certain set of bytes. That said, I'm happy to see improvements in
> > >>this area but I'm flat out upset about the notion that we must be
> > >>perfect here- our checksums themselves aren't perfect for catching
> > >>corruption either.
> > >
> > >The point is that we're not detecting errors that we can detect when
> > >read outside of basebackup. I really entirely completely fail how that
> > >can be defended.
> > >
> > >I think we're making promises with this the basebackup feature we're not
> > >even remotely keeping. I don't understand how you can defend that, given
> > >the current state, you can have a basebackup that you took with
> > >checksums enabled, and then when actually use that basebackup you get
> > >checksum failures. Like it's one thing not to detect all storage
> > >issues, but if we do detect them after using the basebackup, that's
> > >really not ok.
> >
> > Yeah, if basebackup completes without reporting any invalid checksums, but
> > running pg_verify_checksums on the same backups detects those, that probably
> > should raise some eyebrows.
>
> That isn't actually what would happen at this point, just so we're
> clear. What Andres is talking about is a solution which would only
> actually work for pg_basebackup, and not for pg_verify_checksums
> (without some serious changes which make it connect to the running
> server and run various functions to perform the locking that he's
> proposing pg_basebackup do...).

Well, I still think it's just plain wrong to do online checksum
verification outside of the server, and we should just reject adding
that as a feature.

Besides the fact that I think having at precisely equal or more error
detection capabilities than the backend, I think all the LSN based
approaches also have the issue that they'll prevent us from using them
on non WAL logged data. There's ongoing work to move SLRUs into the
backend allowing them to be checksummed (Shawn Debnath is IIRC planning
to propose a patch for v13), and we also really should offer to also
checksum unlogged tables (and temp tables?) - just because they'd be
gone after a crash, imo doesn't make it OK to not detect corrupted on
disk data outside of a crash. For those things we won't necessarily
have LSNs that we can conveniently can associate with those buffers -
making LSN based logic harder.

> I outlined a couple of other approaches to improving that situation,
> which would be able to be used with pg_verify_checksums without having
> to connect to the backend, but I'll note that those were completely
> ignored, leading me to believe that there's really not much more to
> discuss here since other ideas are just not open to being considered.

Well, given that we can do an accurate determination without too much
code in the basebackup case, I don't see what your proposals gain over
that? That's why I didn't comment on them. I'm focusing on the
basebackup case, over the online checksum case, because it's released
code.

Greetings,

Andres Freund