Обсуждение: Fwd: Administrative Processing for csitech: getrows

I have several databases in my instance/cluster.  When I VACUUM ANALYZE
this database, it gives me the following error.  All other database seem
fine.  This looks like I am being told that some of our shared memory parms
are not high enough, yet the other databases in the same instance do not
complain.

We are running postgres 7.1.2 on a E3500 using Solaris 2.6.


>Before Vacuum
>Sat Feb  9 19:00:34 MST 2002
>psql:/usr/local/postgres/etc/csitech/daily/sql/getrows.sql:42:
>NOTICE:  RegisterSharedInvalid: SI buffer overflow
>psql:/usr/local/postgres/etc/csitech/daily/sql/getrows.sql:42:
>NOTICE:  InvalidateSharedInvalid: cache state reset
>VACUUM
>Sat Feb  9 19:00:36 MST 2002
>After Vacuum

--
Naomi Walker
Chief Information Officer
Eldorado Computing, Inc.
602-604-3100  ext 242

Naomi Walker <nwalker@eldocomp.com> writes:
>> psql:/usr/local/postgres/etc/csitech/daily/sql/getrows.sql:42:
>> NOTICE:  RegisterSharedInvalid: SI buffer overflow
>> psql:/usr/local/postgres/etc/csitech/daily/sql/getrows.sql:42:
>> NOTICE:  InvalidateSharedInvalid: cache state reset

These are not errors and can be ignored.  (These messages have been
downgraded to DEBUG level in 7.2.)

            regards, tom lane