Re: [PoC] configurable out of disk space elog level

On Thu, 17 Nov 2022 at 14:56, Robert Haas <robertmhaas(at)gmail(dot)com> wrote:
>
> Having a switch for one particular kind of error (out of many that
> could possibly occur) that triggers one particular coping strategy
> (out of many that could possibly be used) seems far too specific a
> thing to add as a core feature. And even if we had something much more
> general, I'm not sure why that should go into the database rather than
> being implemented outside it. After all, nothing at all prevents the
> user from scanning the database logs for "out of space" errors and
> shutting down the database if any are found. While you're at it, you
> could make your monitoring script also check the free space on the
> relevant partition using statfs() and page somebody if the utilization
> goes above 95% or whatever threshold you like, which would probably
> avoid service outages much more effectively than $SUBJECT.

I have often thought we report a lot of errors to the user as
transaction errors that database admins are often going to feel they
would rather treat as system-wide errors. Often the error the user
sees seems like a very low level error with no context that they can't
do anythign about. This seems to be an example of that.

I don't really have a good solution for it but I do think most users
would rather deal with these errors at a higher level than individual
queries from individual clients. Out of disk space, hardware errors,
out of memory, etc they would rather handle in one centralized place
as a global condition. You can work around that with
middleware/libraries/monitoring but it's kind of working around the
database giving you the information at the wrong time and place for
your needs.