----- Forwarded message from Kai Moritz <k...@ich-geh-kaputt.de> -----
X-Spam-Status: No, score=-2.6 required=5.0 tests=BAYES_00 autolearn=ham
version=3.2.5
Subject: Re: 545...@bugs.debian.org
From: Kai Moritz <k...@ich-geh-kaputt.de>
To: Tiago Bortoletto Vaz <ti...@debian-ba.org>
Date: Wed, 23 Sep 2009 09:33:09 +0200
X-Mailer: Evolution 2.26.1
Am Sonntag, den 20.09.2009, 21:42 -0300 schrieb Tiago Bortoletto Vaz:
Hi!
> I think the issue you reported is not a bug in apticron. Please help me to
> confirm this:
>
> - for some reason the apt-get command called by apticron freezes in your
> system. It would happen anyway if you use another way to call it. I suggest
> you to find the reason which makes 'apt-get -qq update' freeze in your
> system (maybe a network issue?).
No.It did not freeze on any of the systems, which had shown the error.
While apticron was freezed, the tool aptitude correctly refused to work,
because there was another instance running (the one started by apticron)
-- because of that, I discovered the error in the first place. After I
killed the 'apt-get -qq update' all (including apticron and aptitude)
were running fine again on any of the three serverers, which had shown
the error.
>
> - the part of manpage you mentioned in summary says it's not a good idea using
> '-qq' to perform a real action in apt-get because by using this, apt-get
> wont
> ask you to confirm that action. However, it's exactly what we need in
> apticron: updating the sources lists without asking you to confirm :)
>
Yeah, and that is, why I use and love apticron :)
I guess, that something unexpected happend, while 'apt-get -qq update'
was running, which made an user-interaction necessary, so that apt-get
gone to sleep, waiting for the user-response, which it could not
receive.
For example, the mirror-server might habe been updated, while the
command was running (all three servers are using the same mirror-server,
managed by my provider).
> I would like to know how often it's happening in your system. As I said, I'm
> pretty sure it's not a bug in apticron, but I can try to help you there anyway
> if you give me more info. Also, if you agree with me please tell me so I close
> this report.
I apreciate the help, your offering, but the strange behavior described
in the bug-report happend only once. And after killing the sleeping
'apt-get -qq update' all was running fine again. Since then, I have not
seen that strange behavior again.
So my best guess is, that some very unlikely cicumstances (perhaps a
'racecondition' while fetching the update informations from the mirror,
which lead to inconsitent package-informations ?!?) forced apt-get to
ask the user, instead of failing silently.
But if there is really no one else out there, who witnessed a simmilare
behavior of apticron, it might be just a weired sort of misconfiguration
on my side, that shows up only in very special situations. I don't know.
What made me hesitate, so that I filled in the bug-report was the fact,
that exactly the same error happend on three servers. All of them are
managed by me, so I might have made the same configuration-mistake three
times. But the three servers are no clones. Everyone was set up
separatedly not all at the same time.
Greetings
Kai
----- End forwarded message -----
--
Tiago Bortoletto Vaz
http://tiagovaz.org
0xA504FECA - http://pgp.mit.edu
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
