Re: Fwd: Re: A new look at old NFS readdir() problems?

On 1/8/25 12:40, Bruce Momjian wrote:
> On Tue, Jan 7, 2025 at 01:03:05AM +0000, David Steele wrote:
>>>> I'm more concerned about the report we saw on SUSE/NFS [1]. If that
>>>> report is accurate it indicates this may not be something we can just
>>>> document and move on from -- unless we are willing to entirely drop
>>>> support for NFS.
>>>> [1] https://github.com/pgbackrest/pgbackrest/issues/1423
>>>
>>> I installed an up-to-date OpenSUSE image (Leap 15.6) and it passes
>>> my "rmtree" test just fine with my NAS. The report you cite
>>> doesn't have any details on what the NFS server was, but I'd be
>>> inclined to guess that that server's filesystem lacked support
>>> for stable NFS cookies.
>>
>> The internal report we received might have had a similar cause. Sure seems
>> like a minefield for any user trying to figure out if their setup is
>> compliant, though. In many setups (especially production) a drop database is
>> rare.
>
> Will people now always get a clear error on failure?

The error will be something like "directory is not empty" when trying to
drop a database. So not very clear at all.

> Crazy idea, but
> could we have initdb or postmaster start test this?

I think this test would be too expensive for postmaster start as it
would require creating and deleting a large number of files (maybe
multiple times). I suppose it could be a special setting but I doubt
that would be popular.

initdb might be more promising but clusters are frequently
copied/restored to different storage so I'm not sure if it would be too
helpful in the long run.

Regards,
-David