[389-users] high-rate queries

Mon Mar 28 17:04:55 UTC 2011

On 03/28/2011 10:20 AM, Karoly Czovek wrote:
> Let me wait for the next errors, switched back accesslog
> pkg info:
>
>
> Name       : 389-ds-base
> Arch       : i386
> Version    : 1.2.7.5
> Release    : 1.el5
> Size       : 5.4 M
> Repo       : installed
> Summary    : 389 Directory Server (base)
Ok.  I think this is https://bugzilla.redhat.com/show_bug.cgi?id=668619 
which has been fixed in 389-ds-base-1.2.8.a2 - 389-ds-base-1.2.8.rc2 is 
now in testing.  You could see if you can reproduce the problem in a 
non-production environment
yum install --enablerepo=epel-testing 389-ds-base
>
>
>
> On Mar 28, 2011, at 6:11 PM, Rich Megginson wrote:
>
>> On 03/28/2011 10:09 AM, Karoly Czovek wrote:
>>> lance:~# /usr/bin/ldapsearch  -x -LLL uid=karoly.czovek
>>> ldap_result: Can't contact LDAP server (-1)
>>> lance:~# echo $?
>>> 254
>>>
>>>
>>> [here i restarted 389ds on the ldap server]
>>>
>>> lance:~# /usr/bin/ldapsearch  -x -LLL uid=karoly.czovek
>>> dn: uid=karoly.czovek,ou=People,dc=moveone,dc=info
>>> gidNumber: 1010
>>> mailAccountStatus: Active
>>> mailUserHome: /home/vmail/karoly.czovek
>>> [...]
>> Ok.  Thanks.
>> What version of 389-ds-base are you using?  What do you see in the 
>> access log from around the time of a failed client connection 
>> attempt, from ldapsearch or from some other client?
>>>
>>>
>>>
>>>
>>> On Mar 28, 2011, at 5:18 PM, Rich Megginson wrote:
>>>
>>>> On 03/28/2011 09:09 AM, Karoly Czovek wrote:
>>>>> i can reproduce, but the server is in production.
>>>>> if the clients got this error then i also unable to connect with 
>>>>> ldapsearch.
>>>> Ok, what is the error code?
>>>>>
>>>>> just now appeared, i have full verbose trace logs, but nothin 
>>>>> intrestring,
>>>>> but a ton from theese
>>>>>
>>>>> 28/Mar/2011:10:33:00 -0400] - => 
>>>>> slapi_reslimit_get_integer_limit() conn=0xb023ffa8, handle=3
>>>>> [28/Mar/2011:10:33:00 -0400] - <= 
>>>>> slapi_reslimit_get_integer_limit() returning NO VALUE
>>>>> [28/Mar/2011:10:33:00 -0400] - => 
>>>>> slapi_reslimit_get_integer_limit() conn=0xb023fed8, handle=3
>>>>> [28/Mar/2011:10:33:00 -0400] - <= 
>>>>> slapi_reslimit_get_integer_limit() returning NO VALUE
>>>>> [28/Mar/2011:10:33:00 -0400] - => 
>>>>> slapi_reslimit_get_integer_limit() conn=0xb023fe08, handle=3
>>>>> [28/Mar/2011:10:33:00 -0400] - <= 
>>>>> slapi_reslimit_get_integer_limit() returning NO VALUE
>>>>> [28/Mar/2011:10:33:00 -0400] - => 
>>>>> slapi_reslimit_get_integer_limit() conn=0xb023fd38, handle=3
>>>>> [28/Mar/2011:10:33:00 -0400] - <= 
>>>>> slapi_reslimit_get_integer_limit() returning NO VALUE
>>>>> [28/Mar/2011:10:33:00 -0400] - => 
>>>>> slapi_reslimit_get_integer_limit() conn=0xb023fc68, handle=3
>>>> What error log level are you using?  These messages are ok.  The 
>>>> errors log messages are going to be unhelpful at best, and 
>>>> misleading at worst, if you use specific log levels like CONN and 
>>>> TRACE, without looking at the code.
>>>>>
>>>>>
>>>>> for the period. increased the files and restarted the srv, let me 
>>>>> see what happens...
>>>>> any other idea, advice what settings are  needed for large amount 
>>>>> of queries from 389ds  ldap?
>>>> I still don't have enough information from you to make that 
>>>> determination.  What version of 389-ds-base are you using?  What do 
>>>> you see in the access log from around the time of a failed client 
>>>> connection attempt, from ldapsearch or from some other client?
>>>>>
>>>>>
>>>>>
>>>>> On Mar 28, 2011, at 4:42 PM, Rich Megginson wrote:
>>>>>
>>>>>> On 03/28/2011 08:29 AM, Karoly Czovek wrote:
>>>>>>> the server is 32bit, the exact client error message is cannot 
>>>>>>> connect to the server.
>>>>>> But what is the error code?  80?  81?  91?  If you can reproduce 
>>>>>> with ldapsearch, in the shell, $? should hold the error code
>>>>>>> (even ldapsearch cant). let me check the logs once appearing the 
>>>>>>> errors again,
>>>>>>> now disabled the restart cronjob.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Mar 28, 2011, at 4:16 PM, Rich Megginson wrote:
>>>>>>>
>>>>>>>> On 03/28/2011 03:45 AM, Karoly Czovek wrote:
>>>>>>>>> Hi All,
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Is there any reason, workaround for the 389DS about high-rate 
>>>>>>>>> queries?
>>>>>>>>> Acting as LDAP server for one exim and 5 postfix servers, and 
>>>>>>>>> for 6 courier-imap servers, libnss-pam
>>>>>>>> What version of 389-ds-base?  32-bit or 64-bit?
>>>>>>>>>
>>>>>>>>> After a time, my clients cannot connect to the 389DS, ldap 
>>>>>>>>> queries coming back with temporary lookup faulure.
>>>>>>>> What are the exact client error codes/messages?  Any errors in 
>>>>>>>> the dirsrv errors log?  Can you post excerpts of the dirsrv 
>>>>>>>> access log from around the time of the client errors?
>>>>>>>>> The 389DS has 2 cores with 2GB of ram, related settigns are 
>>>>>>>>> the following:
>>>>>>>>>
>>>>>>>>>
>>>>>>>>> Karoly CZOVEK
>>>>>>>>> Global Systems Administrator
>>>>>>>>> MoveOne IT Department
>>>>>>>>>
>>>>>>>>> Eastern Europe - Balkans - CIS&  Central Asia - Middle East& 
>>>>>>>>>  Africa -
>>>>>>>>> Asia Pacific
>>>>>>>>> phone: +36 1 266 0181 - ext.6710
>>>>>>>>> mobile: +36 70 708 9953
>>>>>>>>> skype: mo_karoly.czovek
>>>>>>>>> email: karoly.czovek at moveoneinc.com 
>>>>>>>>> <mailto:karoly.czovek at moveoneinc.com>
>>>>>>>>> web: http://www.moveoneinc.com <http://www.moveoneinc.com/>
>>>>>>>
>>>>>>> -- 
>>>>>>> Karoly CZOVEK
>>>>>>> Global Systems Administrator
>>>>>>> MoveOne IT Department
>>>>>>>
>>>>>>> Eastern Europe - Balkans - CIS&  Central Asia - Middle East& 
>>>>>>>  Africa -
>>>>>>> Asia Pacific
>>>>>>> phone: +36 1 266 0181 - ext.6710
>>>>>>> mobile: +36 70 708 9953
>>>>>>> skype: mo_karoly.czovek
>>>>>>> email: karoly.czovek at moveoneinc.com 
>>>>>>> <mailto:karoly.czovek at moveoneinc.com>
>>>>>>> web: http://www.moveoneinc.com <http://www.moveoneinc.com/>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> 389 users mailing list
>>>>>>> 389-users at lists.fedoraproject.org
>>>>>>> https://admin.fedoraproject.org/mailman/listinfo/389-users
>>>>>>
>>>>>
>>>>> -- 
>>>>> Karoly CZOVEK
>>>>> Global Systems Administrator
>>>>> MoveOne IT Department
>>>>>
>>>>> Eastern Europe - Balkans - CIS&  Central Asia - Middle East&  Africa -
>>>>> Asia Pacific
>>>>> phone: +36 1 266 0181 - ext.6710
>>>>> mobile: +36 70 708 9953
>>>>> skype: mo_karoly.czovek
>>>>> email: karoly.czovek at moveoneinc.com 
>>>>> <mailto:karoly.czovek at moveoneinc.com>
>>>>> web: http://www.moveoneinc.com <http://www.moveoneinc.com/>
>>>>>
>>>>>
>>>>>
>>>>>
>>>>> --
>>>>> 389 users mailing list
>>>>> 389-users at lists.fedoraproject.org
>>>>> https://admin.fedoraproject.org/mailman/listinfo/389-users
>>>>
>>>
>>> -- 
>>> Karoly CZOVEK
>>> Global Systems Administrator
>>> MoveOne IT Department
>>>
>>> Eastern Europe - Balkans - CIS&  Central Asia - Middle East&  Africa -
>>> Asia Pacific
>>> phone: +36 1 266 0181 - ext.6710
>>> mobile: +36 70 708 9953
>>> skype: mo_karoly.czovek
>>> email: karoly.czovek at moveoneinc.com 
>>> <mailto:karoly.czovek at moveoneinc.com>
>>> web: http://www.moveoneinc.com <http://www.moveoneinc.com/>
>>>
>>>
>>>
>>>
>>> --
>>> 389 users mailing list
>>> 389-users at lists.fedoraproject.org
>>> https://admin.fedoraproject.org/mailman/listinfo/389-users
>>
>
> -- 
> Karoly CZOVEK
> Global Systems Administrator
> MoveOne IT Department
>
> Eastern Europe - Balkans - CIS&  Central Asia - Middle East&  Africa -
> Asia Pacific
> phone: +36 1 266 0181 - ext.6710
> mobile: +36 70 708 9953
> skype: mo_karoly.czovek
> email: karoly.czovek at moveoneinc.com <mailto:karoly.czovek at moveoneinc.com>
> web: http://www.moveoneinc.com <http://www.moveoneinc.com/>
>
>
>
>
> --
> 389 users mailing list
> 389-users at lists.fedoraproject.org
> https://admin.fedoraproject.org/mailman/listinfo/389-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.fedoraproject.org/pipermail/389-users/attachments/20110328/d1114076/attachment.html>