Segmentation fault (SEGV) in cyrus ptloader / CentOS 7

[Cornelius Hald]

Thanks for the info Jan. It is not that easy to find good information
about ptloader and Kolab :)

Cheers,
Conny

On Thu, 2015-01-15 at 10:58 +0100, Jan Kowalsky wrote:
> Hi Cornelius,
> 
> 
> Am 12.01.2015 um 12:29 schrieb Cornelius Hald:
> > On Fri, 2015-01-09 at 22:20 +0100, Jan Kowalsky wrote:
> >> Hi all,
> >>
> >> Am 02.01.2015 um 10:33 schrieb Cornelius Hald:
> >>> Hi Franz,
> >>>
> >>> I'm not having any update on that. Also there seems to be no activity on
> >>> the bugs if filed. What I find strange is that there are not more
> >>> complains about that. In my opinion crashes like that should be treated
> >>> with very priority - but maybe it is hare to reproduce?!
> >>
> >> just for information: yes I do complain. I just disabled ptloader for
> >> the moment because it doesn't work for me neither.
> > 
> > Are there any side effects if you disable ptloader completely?
> 
> This was the suggested approach for the multidomain-setup before cyrus 2.5
> 
>   https://docs.kolab.org/howtos/multi-domain.html
> 
> but we got the problem that then cyrus is unable to proof the canonical
> names of users if one is adding a new acl.
> 
> So users could add acls (via roundcube or mailclient) for user subjects
> into cyrus db which don't exist. In the worst case your cyrus db is
> corrupted.
> 
> I wrote a bug-report about this behaviour:
> 
> https://issues.kolab.org/show_bug.cgi?id=3707
> 
> Maybe there are other impacts if canonification of user subjects don't work.
> 
> Best Regards
> Jan