Re: Errors - Mailing list pgsql-general
From | Nigel J. Andrews |
---|---|
Subject | Re: Errors |
Date | |
Msg-id | Pine.LNX.4.21.0204222151240.2455-100000@ponder.fairway2k.co.uk Whole thread Raw |
In response to | Errors (J Norfleet <jnorfleet@picusnet.com>) |
Responses |
PgAccess patching
|
List | pgsql-general |
On Mon, 22 Apr 2002, J Norfleet wrote: > On Monday 22 April 2002 3:15, you wrote: > > On Mon, 22 Apr 2002, J Norfleet wrote: > > > Not sure how general this is but when trying to connect to the server via > > > either a remote application or even a local php script, I get the > > > following error. (It was working?) > > > > > > ERROR: No such attribute or function 'fio' > > > pg_recvbuf: unexpected EOF on client connection > > > > > > Any help is much appreciated. > > > Thanks > > > > What changed or happened? > > Not sure. This project was handed to me a couple of months back and have had > to learn sql as such. Not counting trying to document everything else. > > > > > What version of postgres are you running? What is the host system (of less > > interest)? > > postgressql 7.0.2. Running on slackware 7. > > > > > What happens when you try connecting with psql? > > It connects fine. I can open all of the tables and edit them with no problem. > > > Try increasing the debug level of the postmaster, see postgres.conf file in > > the data directory, and see what messages appears. Though the answer to the > > first question is presumably going to be a good starting place since > > something must have changed or happened. > > I have upped the error reporting to 5 and the fio error is the only report > that seemed relevant. granted I'm still learning what the rest of it means :) > > I was told to try "grant all on <table> to nobody" but that still didn't > help... Ok, so does it matter what your remote client is trying to do? As in have all your remote/PHP tests attempted to use something called fio? (Simple questions sometimes lead in the direction of the problem) If so does the attribute or function exists where your programs expect it to be? It may be the function or column has been dropped. Something may be causing your clients to exit, for example a SEGV, since the message is definitely from the backend reporting an unexpected loss of connection (I'm assuming that the client/server protocol has something in it to 'manage' a clean closing of a connection). Are your clients using the correct version of the libraries etc.? Do the local PHP tests use the same libraries as the local psql? Do you get error reports in the clients' logs? Do you get any core files from the clients? I'm not sure what else to suggest at the moment. I've scanned the 7.0.2 source tree and the only occurance of fio as a string is where the configure scripts are checking for the sfio (safe fast IO) library. Everyone else feel free to chime in :) -- Nigel J. Andrews Director --- Logictree Systems Limited Computer Consultants
pgsql-general by date: