[Kolab-devel] SOLVED: Re: Fwd: Re: can't connect to Sieve -- bug ?
johannes_graumann at web.de
Fri Dec 13 09:18:04 CET 2013
Johannes Graumann wrote:
> Geoff Nordli wrote:
>> -------- Original Message --------
>> Subject: Re: can't connect to Sieve
>> Date: Thu, 21 Nov 2013 12:12:35 -0800
>> From: Geoff Nordli <geoffn at gnaa.net>
>> Reply-To: geoffn at gnaa.net
>> To: users at lists.kolab.org
>> On 13-11-21 12:02 PM, Troy Carpenter wrote:
>>> I had something similar recently. I was able to connect to the daemon,
>>> Kolab and Roundcube had issues. From memory, Kolab said it couldn't
>>> retrieve the scripts, while Roundcube complained about not connecting to
>>> the server.
>>> 1. Check to make sure the /var/lib/imap/sieve directory exists and has
>>> what you would expect below that ("domain/<letter>/<domainname>").
>>> 2. Check permissions (on my system that's 700) and ownership
>> Thanks Troy.
>> You are right, it doesn't have that folder structure under sieve
>> Maybe it is a bug then?
>> Anyone have ideas on when this should be created?
>> I am running CentOS 6.4 with Kolab Enterprise.
>> It seems the Sieve folder doesn't get correctly setup.
>> Can anyone point me to the script that is supposed to configure the
> I have exactly the same configuration and problem - it's the last thing
> that's preventing me from going life. Any pointers on how to fix the
> inexistent sieve structure?
> Geoff, are you running by any chance in an lxc container as well and may
> this be related?
With significant help from the IRC channel i figured out the basis of my
problem: "yum install kolab" had pulled in the EPEL, rather than kolab
version of roundcube. Scrapping the installation and reinstalling with
"priority = 0" inserted for all kolab-related repositories in
/etc/yum.repo.d brought in the correct version and made the issue go away.
As I'm operating in an lxc container scrapping was easy and I did not
investigate fixing the installation in place.
More information about the devel