reassign 629984 fam
retitle 629984 famd can steal ports of other applications, breaking them
severity 629984 grave
thanks
Severity set to grave as famd can break other applications (obviously
seen here), randomly, making things hard to debug.
On 2011-06-09 22:19:32 -0600, Bob Proulx wrote:
> Vincent Lefevre wrote:
> > The problem comes from the fact that famd chooses a random port at
> > startup.
>
> Just to be clear are you saying that famd picked port 783 randomly at
> startup?
Yes, the famd man page says:
-P port
Bind to the specified TCP port instead of choosing at random.
So, by default, the port is chosen at random, and according to
netstat, it was 783 at the last reboot, explaining why spamd
cannot start.
> > I don't know what is the policy for port selection, but either
> > spamd should choose a different port until it succeeds, or (if
> > ports are supposed to be fixed) this should be seen as a critical
> > bug of famd, because its choice potentially breaks other packages
> > (like spamassassin).
>
> Look in /etc/services and you will find that port 783 is allocated in
> Debian to spamd. That is the port that spamc will use to contact the
> spamd. It has been coordinated.
So, this is a bug in fam, and I think that it should be fixed in one
of the following way:
1. start famd with some fixed port with the -P option (and allocate
it), or
2. still allow a random port but not those that are allocated in
/etc/services,
or both.
(1) can simply be done by adding the option in the /etc/init.d/fam
script and doing the allocation (even though the user can modify
/etc/init.d/fam, as being a config file, he cannot do the allocation,
which is quite important, otherwise some future official change to
the allocations can break his config; and anyway Debian's default
config should not do bad things by default).
(2) would improve robustness if for some reason the user wants to
start famd manually without the -P option. Of course, it could be
done in addition to (1).
--
Vincent Lefèvre <vinc...@vinc17.net> - Web: <http://www.vinc17.net/>
100% accessible validated (X)HTML - Blog: <http://www.vinc17.net/blog/>
Work: CR INRIA - computer arithmetic / Arénaire project (LIP, ENS-Lyon)
--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
