Carpenter, Troy | 26 Mar 21:26 2015

Kolab Development Repository

Since I'm on a roll in sending emails to the list today, I thought I'd 

On my test system, I recently upgraded up to 3.4 in Centos 6, then 
activated the development repositories right afterward.  The only 
packages that updated were a number of Roundcube packages, and 
cyrus-imapd from 2.4 to 2.5, which is what I kind of expected right 
after 3.4 was released...that the development packages haven't moved 
much beyond 3.4 yet.  Apart from switching to nightly builds, is using 
the development repository the best way to stay closer to the edge (with 
all the caveats that go with using development packages, of course)?

I noticed that with cyrus-imap at 2.5, I didn't see all my folders in 
Roundcube, but only saw the Inbox.  When I downgraded back to 2.4 (from 
the 3.4 repository), then all the folders reappeared.  Is there a thread 
about that somewhere else?  Is there a config change I need to make in 
order to upgrade, or is it a bug that hasn't been fixed yet?



Troy Carpenter
troy at carpenter dot cx
Tobias Brunner | 26 Mar 15:29 2015

Cleanup duplicate contacts


After the upgrade troubles (3.3->3.4) with the missing roundcube/libkolab db 
schema upgrades I'm now having a lot of duplicate contacts (many contacts are 
even there several times). Is there a way to automatically cleanup this mess? 
Doing it by hand is very cumbersome =(

Thanks for any inputs

Michael Schumacher | 26 Mar 10:37 2015

strange problem connecting Outlook 2013 and Kolab 3.4 (ActiveSync)

good morning,

I installad the most recent version of Kolab (3.4) on a centos7 server
and everything seems to work fine.
The only problem I am facing is when I try to connect an Outlook2013
client via ActiveSync.

The machine is part of my private
( company network, the machine is in
the "public" at a root server provider.

I can login to and get the
"it works" response.
I can "nslookup" successfully in my private
( network from the windows client.

When I try to setup outlook using this
howto, Outlook cannot resolve the IP-address correctly. Instead, it
connects to my webserver and complains about an
outdated certificate. The certificate is actually expired.

"nslookup" and "nslookup" both resolve to
the public IP-address.

The puzzling thing is that outlook does not seem to resolve the mail
server name correctly to my internal machine,
it resolves the to the machine instead.

(Continue reading)

Markelov Alexander | 26 Mar 09:55 2015

IMAP folders subscription


I'm trying to integrate Outlook 2010 as IMAP-only client to Kolab and 
encountered a issue.
Outlook-IMAP-user subscribed to all unnecessary folders, like - Archive, 
Calendar, Configuration, Contacts, etc...
Cyrus 2.5 has an option - autocreate_subscribe_folders. But if I setup 
autocreate_subscribe_folders: Sent|Trash, Kolab is still subscribe to 
all folders.
Can I disable subscription to all Kolab folders?


Markelov Alexander - IT/Net engineer
xmpp:deep <at>
ICQ 13372829
Markelov Alexander | 25 Mar 10:08 2015

Kolab 3.4 on Debian Wheezy experience.

Some experience about Kolab 3.4 on Debian Wheezy after upgrade from 
Kolab 3.3

1. You need to look after for kolab_cache_* tables in the roundcube DB. 
Sometimes they grow!  If something wrong - use 
2. Best practice is use option - "innodb_file_per_table" in MySQL. If 
something go wrong with kolab_cache_*, tables only 
"innodb_file_per_table" option allow You to reclaim disk space.
3. PHP garbage collector switched off in Debian by default, and maybe is 

cat /etc/php5/apache2/php.ini | grep session.gc_probability
session.gc_probability = 0

You need to think about it and clean garbage manually like:

cat /etc/cron.daily/rc-session-clean
mysql -v -v --user=roundcube --password=roundcubepassword 
--database=roundcube --execute="DELETE FROM session WHERE changed < 

First clean the roundcube.session table.
Second clean roundcube.cache_* tables.
(Continue reading)

Stelios A. | 25 Mar 09:29 2015

kolab_cache_contact huge


I've truncate a while ago the kolab_cache_contact and kolab_cache_event table and today after a few weeks checked and its size is 48GB and 30GB for the event table.

This is a very low volume email server (getting about 150 emails per day) and honestly have no idea why those tables are getting huge.

Is it safe to just add a cron with a script to truncate them?


users mailing list
users <at>
dsp3 | 24 Mar 18:19 2015

Re: EPEL upgrade of Kolab roundcubemail?

Take a look at this:

On 24/03/15 16:33, ybk <at> wrote:
> I asked about this once already, but suggested response didn't work.
> How exactly yum configuration should be modified?
> On Thursday, March 19, 2015 12:47 PM, dsp3 <info <at>> wrote:
>> You need to tell yum to prioritize Kolab repo over EPL repo.
>> On 19/03/15 14:45, Cornelius Hald wrote:
>>> Hi,
>>> for some time now I have a pending update of 'roundcubemail'. It always
>>> looked odd because it wanted to upgrade from 1.1 to 1.1.0.
>>> So today I've looked at it closer and it seems like yum wants to
>>> install
>>> the vanilla roundcube version from EPEL and replace the Kolab roundcube
>>> version by doing so.
>>> Shouldn't the Kolab version of Roundcube get a different name like
>>> kolab-roundcube to prevent something like this?
>>> Or do I misunderstand something? Here is the output of "yum info
>>> roundcubemail".
>>> Installed Packages
>>> Name        : roundcubemail
>>> Arch        : noarch
>>> Version     : 1.1
>>> Release     : 20.9.el7.kolab_3.3
>>> Size        : 14 M
>>> Repo        : installed
>>> >From repo   : Kolab_3.3_Updates
>>> [...]
>>> Available Packages
>>> Name        : roundcubemail
>>> Arch        : noarch
>>> Version     : 1.1.0
>>> Release     : 1.el7
>>> Size        : 2.8 M
>>> Repo        : epel/x86_64
>>> [...]
>>> Cheers,
>>> Conny
>>> _______________________________________________
>>> users mailing list
>>> users <at>
>> _______________________________________________
>> users mailing list
>> users <at>
Herwig Unterrichter | 24 Mar 14:37 2015

Kolab 3.4 - how to delete distribution group in Global Adress Book


After deleting a Kolab Distribution Group using Kolab-Webadmin, the
group does not get deleted in the Global Adress Book.

Do I have to manually trigger anything or do I have to manually delete
the group? And how do i do that?

Thanks for any help,

Herwig Unterrichter | 24 Mar 00:36 2015

Kolab 3.4 - change default UID and primary-email-address

Hi Everybody,

I would like to change the settings for the automatic creation of the
UID and primary-email-address when I add a user in the kolab-webadmin.

I would like to have the UID consist of "givenname.surname" instead of
only "surname"

Does somebody know how to achieve that?

Thanks for any help.

Michael Grunewalder | 23 Mar 14:08 2015

invitations as attachments on iOS


when i receive meeting request or appointments, they are delivered as attachments to my email address,
they don’t arrivee in the inbox of my iOS calendar. Last version checked was iOS 8.2. Would anyone know
how to fix that?
It works perfectly fine in Zarafa, and i think both, Zarafa and Kolab use z-push, right?

Thanks a lot in advance

users mailing list
users <at>
Matthias Busch | 23 Mar 11:59 2015

still... wallace troubles in kolab 3.4

Kolab 3.4 (upgraded from 3.3) on Debian 7.8 x64

On sunday evening I noticed that I did not get mails for a few days. 
Apparently wallace was keeping them.

looking at the /var/log/kolab/pykolab.log I see different messages:

2015-03-22 18:38:39,419 pykolab.wallace WARNING No contents configured 
for footer module
--> only warning, not relevant, but while we are discussing wallace, 
what exactly does it want in the none existent footer files? stuff to 
add to the email?


UnicodeEncodeError: 'ascii' codec can't encode character u'\xe4' in 
position 4: ordinal not in range(128)

lots of them, always the same character. in the footer.execute() 
module... copy and paste:

2015-03-22 18:33:15,412 pykolab.wallace ERROR Module footer.execute() 
failed on message 
'/var/spool/pykolab/wallace/invitationpolicy/incoming/tmpOxKgSQ' with 
error: Traceback (most recent call last):
   File "/usr/lib/python2.7/dist-packages/wallace/", line 65, 
in pickup_message
     result_filepath = modules.execute(module, filepath)
   File "/usr/lib/python2.7/dist-packages/wallace/", line 116, 
in execute
     return modules[name]['function'](*args, **kw)
   File "/usr/lib/python2.7/dist-packages/wallace/", 
line 81, in execute
     exec('modules.cb_action_%s(%r, %r)' % ('ACCEPT','footer', filepath))
   File "<string>", line 1, in <module>
   File "/usr/lib/python2.7/dist-packages/wallace/", line 276, 
in cb_action_ACCEPT
     log.debug(_("recipients: %r") % (recipients))
   File "/usr/lib/python2.7/dist-packages/pykolab/", line 
29, in <lambda>
     _ = lambda x: current.lgettext(x)
   File "/usr/lib/python2.7/", line 366, in lgettext
     return tmsg.encode(locale.getpreferredencoding())
UnicodeEncodeError: 'ascii' codec can't encode character u'\xe4' in 
position 4: ordinal not in range(128)

I remember having had problems with wallace and character encoding 
before. I was told it had been fixed and I thought the fixed wallace 
would find its way into 3.4
Then again, I dont know if that problem is related or if it is a 
different encode error.
HOWEVER, looking at the time I must conclude that those errors still let 
the mail be delivered...


these errors did not:

2015-03-22 18:38:39,397 pykolab.wallace ERROR Module 
invitationpolicy.execute() failed on message 
'/var/spool/pykolab/wallace/resources/incoming/tmpTqfD80' with error: 
Traceback (most recent call last):
   File "/usr/lib/python2.7/dist-packages/wallace/", line 65, 
in pickup_message
     result_filepath = modules.execute(module, filepath)
   File "/usr/lib/python2.7/dist-packages/wallace/", line 116, 
in execute
     return modules[name]['function'](*args, **kw)
line 243, in execute
     os.rename(filepath, new_filepath)
OSError: [Errno 2] No such file or directory

I dont know what went wrong here... I restarted wallace and the missing 
mails were delivered, no other errors were shown (except the missing 
contents for footer module)


so first of all, if there is something I can do via config files, I'll 
gladly try it.
If you need more information, Ill try to supply what you need...

more importantly however, wallace really does offer interesting 
possibilities and I really would miss them. but it causes problems on a 
regular basis (which is why I did not activate it on the second kolab 
server where mails MUST work)

I really hope it gets some love from someone who can give it 
(unfortunately, thats not me) and if it does, will get updated with the 
current kolab release and not held back for the upcoming one :)

from what little I know about python, maybe wallace would benefit from 
errorhandling (try ... except ...) to ensure that if anything breaks at 
all wallace doesnt just quit but runs a function (which can handle 
logging of the error but more importantly) ensures the email is delivered!