On Thu, May 23, 2013 at 04:48:20PM +0200, Jakub Hrozek wrote:
> On Thu, May 23, 2013 at 08:44:52AM -0500, Trent Johnson wrote:
>>
>> On May 23, 2013, at 8:02 AM, Lukas Slebodnik <lslebodn(a)redhat.com> wrote:
>>
>>> On (23/05/13 07:28), Trent Johnson wrote:
>>>> Hi,
>>>>
>>>> After upgrading from rhel6.3 to the 6.4 sssd-1.9.2-82.7.el6_4.x86_64 we
can no longer use enumerate=true with active directory and rfc2307bis.
>>>>
>>>> Upon starting the service, sssd_be consumes 100% of the cpu and getent
takes a long time to return results. In the logs we see:
>>>>
>>>> /var/log/messages sssd[be[TESTWIN]]: dereference processing failed :
Input/output error
>>>> /var/log/sssd/sssd.log [sssd] [mt_svc_sigkill] (0x0010): [TESTWIN][10702]
is not responding to SIGTERM. Sending SIGKILL.
>>>>
>>>> Thanks,
>>>> Trent
>>>>
>>> It is not possible to say, where is the problem from two lines of log.
>>> One enumeration bug was fixed in rhel6.4.
>>>
>>> Do you think, that (
https://bugzilla.redhat.com/show_bug.cgi?id=953165) is
the
>>> same problem as you have?
>>>
>>> LS
>>>
>>
>> I am not authorized to view that bug.
>
> I'm sorry Trent, that bug contains some confidential customer
> information and was marked as private. The corresponding upstream ticket
> is
https://fedorahosted.org/sssd/ticket/1893
>
>> The above was the only thing in the logs. I can turn on debug and send you the
logs off list if you like.
>>
>> Thanks,
>> Trent
>
> If I built you a test package, would you be able to test it? The problem
> was not really easy to spot from the logs.
This might actually be hitting more people, so I built the test packages
and put them into my fedorapeople space:
http://jhrozek.fedorapeople.org/sssd-enum-bug/
Thank you. I have tested this, and it is working well. Will this be
ported into rhel, or should I open a bug there?
Thanks,
Trent