[SSSD] IPA client randomly lose memory of users

Qing Chang qchang at sri.utoronto.ca
Fri Dec 7 17:48:20 UTC 2012


On 2012/12/6 13:54, Jakub Hrozek wrote:
> On Thu, Dec 06, 2012 at 01:49:14PM -0500, Qing Chang wrote:
>> On 06/12/2012 12:12 PM, Jakub Hrozek wrote:
>>> On Thu, Dec 06, 2012 at 11:37:41AM -0500, Qing Chang wrote:
>>>> On 05/12/2012 2:58 PM, Dmitri Pal wrote:
>>>>> On 12/05/2012 02:00 PM, Qing Chang wrote:
>>>>>> On 05/12/2012 12:59 PM, Simo Sorce wrote:
>>>>>>> On Wed, 2012-12-05 at 11:31 -0500, Qing Chang wrote:
>>>>>>>
>>>>>>> I see this:
>>>>>>>
>>>>>>>> First failed "getent passwd shassan" triggers:
>>>>>>>> (Wed Dec  5 10:45:05 2012) [sssd[nss]] [nss_cmd_getpwnam_dp_callback]
>>>>>>>> (0x0040): Unable to get
>>>>>>>> information from Data Provider
>>>>>>>> Error: 1, 11, Fast reply - offline
>>>>>>>> Will try to return what we have in cache
>>>>>>> but later on you report only logs after 10:45:05.
>>>>>>>
>>>>>>> It would be useful to see what happened in the provider immediately
>>>>>>> before the error was returned. In particular I would like to see what
>>>>>>> error caused the provider to go offline and if it mentions removing the
>>>>>>> user in any way.
>>>>>> the problem is we don't know when the loss of memory happened and it
>>>>>> appears to
>>>>>> me that nothing is logged about it. If you don't mind, I can send in
>>>>>> more logs.
>>>>>>>> (Wed Dec  5 10:45:48 2012) [sssd[be[sri.utoronto.ca]]]
>>>>>>>> [sdap_save_user] (0x0400): Storing info for
>>>>>>>> user shassan
>>>>>>>> (Wed Dec  5 10:45:48 2012) [sssd[be[sri.utoronto.ca]]]
>>>>>>>> [sysdb_search_user_by_name] (0x0400): No such
>>>>>>>> entry
>>>>>>>> (Wed Dec  5 10:45:48 2012) [sssd[be[sri.utoronto.ca]]]
>>>>>>>> [sysdb_add_user] (0x0400): Error: 17 (File
>>>>>>>> exists)
>>>>>>>> (Wed Dec  5 10:45:48 2012) [sssd[be[sri.utoronto.ca]]]
>>>>>>>> [sysdb_store_user] (0x0080): A user with the
>>>>>>>> same UID [54184] was removed from the cache
>>>>>>>> (Wed Dec  5 10:45:48 2012) [sssd[be[sri.utoronto.ca]]]
>>>>>>>> [sysdb_search_user_by_uid] (0x0400): No such
>>>>>>>> entry
>>>>>>> This is interesting, do you have duplicated UIDs in your directory by
>>>>>>> chance ?
>>>>>> That was also what I was wondering about, but that's her id, it is
>>>>>> unique.
>>>>> Can you compare the LDAP entry of this user to couple other users that
>>>>> are OK, use SSSD regularly and do not have an issue?
>>>>> May be there is something else in the user entry that would shed the light.
>>>> there is nothing different after comparing with a few accounts. The cmd line I used
>>>> is this:
>>>> ldapsearch -LL -x -D 'cn=Directory Manager' -W -b 'dc=sri,dc=utoronto,dc=ca' '(uid=shassan)'
>>>>
>>>> Could there be still something not showing by this cmd?
>>> The term "uid" is kind of overloaded here. We're looking for non-unique
>>> UID numbers, not the uid LDAP attributes, which store the name.
>>>
>>> Something like this might show the offender:
>>> ldapsearch -LL -x -D 'cn=Directory Manager' -W -b 'dc=sri,dc=utoronto,dc=ca' '(uidNumber=54184)'
>> Thanks Jakub, there is a conflict uidNumber indeed.
>>
>> What I tried was "getent passwd 54184", it gave me just one entry which got me to believe
>> that it is unique...
>>
> Right, the getent functions stop at first match.
>
> Does your setup work fine after removing the duplicate UID?
yes. I have removed the cron job.

Many thanks,

Qing



More information about the sssd-devel mailing list