Yes, so something must have killed the server and shut it down leading to
the stale socket message. However I guess without further checking through
the server logs I won't find out exactly why etc.

11/22/05, J Cameron Cooper <jccooper (AT) enfoldsystems (DOT) comwrote:

michael nt milne wrote:
, but Zope and Apache were serving nothing until I re-started Zope.
The server had died for some reason

Indeed. And that's what it tells you with the "unlinking stale socket":
it was not shut down in an orderly manner previously.

But that message says nothing about why.

11/22/05, *Chris McDonough* <chrism (AT) plope (DOT) com
<mailto:chrism (AT) plope (DOT) com>wrote:

I think it probably was running after the first time you invoked it.
The message is only informative, meaning that zopectl died unexpectly
the last time it was run (without cleaning up after itself). It's
not a fatal error or at least hasn't been whenever I've seen it.

Nov 22, 2005, at 10:57 AM, michael nt milne wrote:

Yes, I noticed that the deamon wasn't running and so re-started it.
It started up fine but that was the message I received. Just
wondered what happened etc

11/22/05, Chris McDonough <chrism (AT) plope (DOT) com
<mailto:chrism (AT) plope (DOT) com>wrote: Note that
Zope likely started. Did it? Did you try zopectl stop and
zopectl start again? If so, did it work?

- C

Nov 22, 2005, at 9:52 AM, michael nt milne wrote:

Hi

My zopeinstance went down and I was getting an 'upstream server'
unknown message from apache. restarting zope from ./zopectl this
was the message I got

unlinking stale socket
/,sleep/
No handlers could be found for logger 'root'

Does anyone have any idea on how to prevent this at all, or what
happened here?

Thanks

Michael
--