OOT: Pyzor discovery problem

[Glenn Steen]

On 15/09/06, Glenn Steen <glenn.steen at gmail.com> wrote:
(snip)
> Might be a file that's corrupted... Try doing the discover (and a
> subsequent ping) to some other directory. If that works, simple remove
> the old and redo, if not, well... then I suppose a reinstall is in
> order (which would rewrite any corrupted files). You shouldn't need
> disable pyzor, bu... better safe than sorry, I suppose:-).
> IIRC doing discoveries all the time (well, once per day:-) isn't
> really necessary with pyzor... Unless I remember wrong there only is
> one server, and thats the way things are. So one could well hold of
> with the discoveries until one sees that pyzor just don't work:-).
>
> A philosophical musing one could have is on the reliability of a
> service that only have one central server ... :-). Puts all those
> "TimoutErrors" into perspective:-)
>
(Yeah yeah... me proving I'm a postfix user, by replying to myself:-)

If this is due to fs corruption, you should (of course) endevor to do
some fsck'ing... Even journalised filesystems (read "ext3") might need
that from time to time...


-- 
-- Glenn
email: glenn < dot > steen < at > gmail < dot > com
work: glenn < dot > steen < at > ap1 < dot > se