Wierd Error on update

Hi All,

I am relative PostGres newbie, so if there is a simple answer to this
question, please feel free to send me a link or explaination.

I am running Postgres 7.3 on a intel Linux Redhat 7.3 base.

I am getting two errors which are a bit confounding.

ERROR: pg_class_aclcheck:relation 474653086 not found

on one table within a large SQL file update (+5000 lines with
documentation) that needs to be run from the command line.

Both statements are multiline Update statements. This is straight
SQL, not function code.
Now I can guareentee that the queries refer to an existing table, but
the table is being rebuilt in the core of the long SQL update.

The table in question now has all permissions (my attempt to get
fixed, but that did not help).

Interesting aspect to this problem is that if your run the large
update in one file, and then run the two update statements in a second
file it works fine.
Now I placed the erroring statements at the end of the large file to
make sure it was not a ordering error, but that didnt help.

Any ideas?
Thank you
Thomas Seeber

---------------------------(end of broadcast)---------------------------
TIP 3: if posting/reading through Usenet, please send an appropriate
subscribe-nomail command to majordomo@postgresql.org so that your
message can get through to the mailing list cleanly

Re: Wierd Error on update

Thomas Seeber writes:
> I am running Postgres 7.3 on a intel Linux Redhat 7.3 base.

Original 7.3 release, or (I hope) 7.3.something?

> I am getting two errors which are a bit confounding.
> ERROR: pg_class_aclcheck:relation 474653086 not found

Are there any views involved? Is the statement invoking any functions?
How about triggers? Foreign keys? Rules?

Also, is it always the same number in the error, or does that change
from run to run?

The problem looks a bit like a corrupted rule or view, but I'm not
entirely sure how one would get involved in a simple UPDATE.

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 4: Don't 'kill -9' the postmaster

Re: Wierd Error on update

> Original 7.3 release, or (I hope) 7.3.something?
Red Hat 7.3.2.96-113
Linux Version 2.4.20-20.7custom from cat /proc/version

> > I am getting two errors which are a bit confounding.
> > ERROR: pg_class_aclcheck:relation 474653086 not found
>
> Are there any views involved?
No views.

Is the statement invoking any functions?
The statement is not invoking any fuctions
However heres the statements,

UPDATE school_info_l SET ind_default_data = 't';
UPDATE school_info_l SET sis_system_id = 9 where school_id IN (492, 434, 436);

which are amazingly simple.
Both queries return the same exact error, same error.

not your
> How about triggers? Foreign keys? Rules?
Neither of the updates touch any rules.
sis_system_id has a foreign key constaint to another table that does
in it contain a value of 9 for sis_system_id in that table, but the
first isnt touching anything that has a trigger or foreign key
constraint attached to the column.

The table school_info_l has many contraints that affect other tables however

> Also, is it always the same number in the error, or does that change
> from run to run?
The number is the same for each time I run the update file on the same
database. with this two queries having the same error.
Same result to, if I run these two after the first update file
finishes there is no problem, but if I try to run it at the end of hte
first file, choke.

Very Odd problem to me, since the update statemenets are so simple.
-Tom

---------------------------(end of broadcast)---------------------------
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to majordomo@postgresql.org)

Re: Wierd Error on update

Thomas Seeber writes:
>>> I am getting two errors which are a bit confounding.
>>> ERROR: pg_class_aclcheck:relation 474653086 not found

>> How about triggers? Foreign keys? Rules?

> Neither of the updates touch any rules.
> sis_system_id has a foreign key constaint to another table that does
> in it contain a value of 9 for sis_system_id in that table, but the
> first isnt touching anything that has a trigger or foreign key
> constraint attached to the column.

> The table school_info_l has many contraints that affect other tables however

I would guess that one or another of these foreign-key constraints is
broken. I'd suggest looking through pg_trigger to see if there are
any entries with tgconstrrelid = 474653086. I wouldn't necessarily
recommend just deleting such an entry, mind you, but finding where the
problem is is the first step...

regards, tom lane

---------------------------(end of broadcast)---------------------------
TIP 1: subscribe and unsubscribe commands go to majordomo@postgresql.org