Re:

On 10/19/2004 11:41 AM, Andrew Sullivan wrote:

> On Tue, Oct 19, 2004 at 01:28:52PM +0100, Ben Osborne wrote:
>> Postgres 7.0.2 Problem
>> -----------------------------------
>
> Yikes. That's old.
>
>> (only) database which is running (other than template1). I suspect that the
>> files in the data directory have been conrrupted or otherwise lost
>> integrity, possibly due to the fact that the disk partition where the db
>> cluster is running has reached 100% usage.
>>
>> The problem is the -l doesn't list my database in the catalog, although i
>> can psql xxxxx into the database. Further, \d lists 'No Relations', however
>> i can select * from a table, and although the database reports no rows, the
>> table scheme does report correctly.
>
> I doubt that's your problem. I suspect xid wraparound instead.
> Unfortunately, without a pg_dump, I suspect your data is
> inaccessible (see the current docs, section 21.1.3, for an
> explanation of why this is. I think the 7.0 docs don't contain all
> that info, BTW). We have a very dangerous tool we've used for testing
> that will thump the xid in 7.2, but I have no idea whether that'd
> work in versions prior to that. Jan Wieck might know, though.
>
> A
>

It's a user defined function in C that modifies this xid counter on
call. But I don't think I would recommend mucking with DDL statements in
a wrapped around DB ... that's not going anywhere.

Jan

--
#======================================================================#
# It's easier to get forgiveness for being wrong than for being right. #
# Let's break this rule - forgive me. #
#================================================== JanWieck(at)Yahoo(dot)com #