[aerogear-dev] AG UPS deadlock

[Matthias Wessendorf]

On Fri, Apr 10, 2015 at 12:32 PM, Ivan Gürtler <ivan.gurtler at ahead-itec.com>
wrote:

> And is it possible in current version to set up time for scheduler not to
> run at midnight?
>

unfortunately not - we discussed that option, but I am overasked if we
actually have a JIRA for that :-)
But I agree, it should be done, for 1.1.x - for now you would have to tweak
your fork :)


>
> *Mgr. Ivan Gürtler*
> Mobile software developer
>
> AHEAD iTec, s.r.o., Botanická 554/68a,
> 602 00 Brno (Czech Republic)
>
> www.ahead-itec.com | twitter <https://www.twitter.com/AHEADiTec> | mobile
> security solutions
>
> 2015-04-10 11:20 GMT+02:00 Matthias Wessendorf <matzew at apache.org>:
>
>> Oh, that's this guy, triggered at midnight:
>>
>> https://github.com/aerogear/aerogear-unifiedpush-server/blob/master/service/src/main/java/org/jboss/aerogear/unifiedpush/service/metrics/DeleteOldPushMessageInformationScheduler.java
>>
>> It's a singleton, but per node :-/
>>
>> Can you create a JIRA ticket for that  ?
>>
>> On Fri, Apr 10, 2015 at 10:33 AM, Ivan Gürtler <
>> ivan.gurtler at ahead-itec.com> wrote:
>>
>>> Hi, we use two AG unified push servers with Oracle DB in cluster, but we
>>> have some problem with deadlock (probably caused by clean up by both
>>> servers at the same time).
>>> We used DB owner PUSHSRVAG and application DB account APUSHSRVAG.
>>>
>>> (version ... my fork (with small change in dev) ...
>>> https://github.com/IvanGurtler/aerogear-unifiedpush-server/tree/MyBranch
>>> )
>>>
>>> Is possible change some settings to prevent deadlock ? or some different
>>> solution ?
>>>
>>>
>>> -----------------------------------------------------------------------------------------
>>>
>>> *** 2015-04-10 00:00:04.506
>>> *** SESSION ID:(84.48937) 2015-04-10 00:00:04.506
>>> *** CLIENT ID:() 2015-04-10 00:00:04.506
>>> *** SERVICE NAME:(SYS$USERS) 2015-04-10 00:00:04.506
>>> *** MODULE NAME:(JDBC Thin Client) 2015-04-10 00:00:04.506
>>> *** CLIENT DRIVER:(jdbcthin) 2015-04-10 00:00:04.506
>>> *** ACTION NAME:() 2015-04-10 00:00:04.506
>>>
>>>
>>>
>>> *** 2015-04-10 00:00:04.506
>>> DEADLOCK DETECTED ( ORA-00060 )
>>> See Note 60.1 at My Oracle Support for Troubleshooting ORA-60 Errors
>>> [Transaction Deadlock]
>>>
>>> The following deadlock is not an ORACLE error. It is a
>>> deadlock due to user error in the design of an application
>>> or from issuing incorrect ad-hoc SQL. The following
>>> information may aid in determining the deadlock:
>>>
>>> Deadlock graph:
>>>
>>>        ---------Blocker(s)--------
>>>  ---------Waiter(s)---------
>>> Resource Name                                              process
>>> session holds waits                process session holds waits
>>> TM-00013891-00000000-00000000-00000000        142      84    SX   SSX
>>>                     144     162    SX   SSX
>>> TM-00013891-00000000-00000000-00000000        144     162    SX   SSX
>>>                     142      84    SX   SSX
>>>
>>>
>>> Rows waited on:
>>>   Session 84: no row
>>>   Session 162: no row
>>>
>>> ----- Information for the OTHER waiting sessions -----
>>> Session 162:
>>>   sid: 162 ser: 54036 audsid: 10790094 user: 115/APUSHSRVAG
>>>     flags: (0x41) USR/- flags_idl: (0x1) BSY/-/-/-/-/-
>>>     flags2: (0x40009) -/-/INC
>>>   pid: 144 O/S info: user: oracle, term: UNKNOWN, ospid: 11534420
>>>     image: oracle at s16000ks
>>>   client details:
>>>     O/S info: user: jbotp, term: unknown, ospid: 1234
>>>     machine: s1.sk.srv.acc.sys program: JDBC Thin Client
>>>     application name: JDBC Thin Client, hash value=2546894660
>>>   current SQL:
>>>   delete from PUSHSRVAG.PushMessageInformation where submitDate<:1
>>>
>>> ----- End of information for the OTHER waiting sessions -----
>>>
>>> Information for THIS session:
>>>
>>> ----- Current SQL Statement for this session (sql_id=bsk109wpq55yg) -----
>>> delete from PUSHSRVAG.PushMessageInformation where submitDate<:1
>>>
>>>
>>> -----------------------------------------------------------------------------------------
>>>
>>>
>>>
>>>
>>> Thanks
>>>
>>>
>>> *Mgr. Ivan Gürtler*
>>> Mobile software developer
>>>
>>> AHEAD iTec, s.r.o., Botanická 554/68a,
>>> 602 00 Brno (Czech Republic)
>>>
>>> www.ahead-itec.com | twitter <https://www.twitter.com/AHEADiTec> |
>>> mobile security solutions
>>>
>>> _______________________________________________
>>> aerogear-dev mailing list
>>> aerogear-dev at lists.jboss.org
>>> https://lists.jboss.org/mailman/listinfo/aerogear-dev
>>>
>>
>>
>>
>> --
>> Matthias Wessendorf
>>
>> blog: http://matthiaswessendorf.wordpress.com/
>> sessions: http://www.slideshare.net/mwessendorf
>> twitter: http://twitter.com/mwessendorf
>>
>> _______________________________________________
>> aerogear-dev mailing list
>> aerogear-dev at lists.jboss.org
>> https://lists.jboss.org/mailman/listinfo/aerogear-dev
>>
>
>
> _______________________________________________
> aerogear-dev mailing list
> aerogear-dev at lists.jboss.org
> https://lists.jboss.org/mailman/listinfo/aerogear-dev
>



-- 
Matthias Wessendorf

blog: http://matthiaswessendorf.wordpress.com/
sessions: http://www.slideshare.net/mwessendorf
twitter: http://twitter.com/mwessendorf
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.jboss.org/pipermail/aerogear-dev/attachments/20150410/de5a30ef/attachment-0001.html