Re: cache lookup failed for function 0

On Sat, 30 Sep 2023 13:20:14 -0400 Tom Lane wrote:

>pf(at)pfortin(dot)com writes:
>> Python script and sample file attached...
>
>This runs fine for me, both in HEAD and 15.4.
>
>(Well, it fails at the last GRANT, seemingly because you wrote
>"{table}" not "{TABLE}". But the COPY goes through fine.)

Sorry, I ripped out some other test code which would have been
unnecessarily confusing; but couldn't check the remaining bits due to the
prior failure...

>Assuming that you verified that this specific test case fails
>for you, we're down to just a couple of theories:
>
>1. There's some relevant info you didn't supply yet (is it
>really a completely plain-vanilla installation? Maybe some
>odd choice of locale?)

As vanilla as it gets... Standard locale (C). The only odd thing that
happened: a system update the other day installed and started something
called tracker-miners which I was not happy with:
https://bugs.mageia.org/show_bug.cgi?id=32340
I was quite upset that the distro guys would install/run what they thought
could be a "cool" tool (as you'll see in the above link).

>2. There's something wrong with your Postgres installation.

Scary; but that's what I'm suspecting too now...

>It's hard to get much further than that with the info that
>we have.

Understood; I wasn't sure which rabbit hole to go down; but this is
starting to smell like a DB issue... I'll see about re-installing the
server and pray the DB is sane... If the above tracker-miners found my
postgres installation and added something to it; then I will be EXTREMELY
upset...

Thanks for confirming my simple SQL runs...

> regards, tom lane

Thanks Tom!