Re: enq: ZZ - update hash tables

From: Tanel Poder <tanel_at_tanelpoder.com>
Date: Fri, 11 Oct 2013 16:53:54 +0300
Message-ID: <CAMHX9JJn9-yuYkWpS_q2G9HOZLf5KYWS8DeUL0gF0UikBD3ZMQ_at_mail.gmail.com>



Indeed, it probably clears global context values. KSBXIC stands for "kernel service background cross-instance call". Never seen this event show up before...

SQL> *_at_lt zz*
Show lock type info from V$LOCK_TYPE for lock zz

TYPE LOCK NAME                      ID1 MEANING               ID2 MEANING
            LT_ DESCRIPTION
---- ------------------------------ -------------------------

------------------------- ---
------------------------------------------------------------ ZZ Global Context Action *KSBXIC Action* 0 NO *Lock held for updating Global context hash tables*

SQL> *_at_sed "enq: ZZ"*
Show wait event descriptions matching %enq: ZZ%..

EVENT# EVENT_NAME                                              WAIT_CLASS
        PARAMETER1                PARAMETER2                PARAMETER3
           ENQUEUE_NAME                   REQ_REASON
REQ_DESCRIPTION
------ -------------------------------------------------------
-------------------- ------------------------- -------------------------

------------------------- ------------------------------
-------------------------------- ---------------------------------------------------------------------------------------------------- 827 enq: ZZ - update hash tables Other name|mode KSBXIC Action 0 Global Context Action update hash tables
lock held for updating global context hash tables
-- 
*Tanel Poder*
Enkitec (The Exadata Experts)
Training <http://blog.tanelpoder.com/seminar/> |
Troubleshooting<http://blog.tanelpoder.com/>
 | Exadata<http://www.amazon.com/Expert-Oracle-Exadata-Apress/dp/1430233923>
 | Voicee App <http://voic.ee/>



On Thu, Oct 10, 2013 at 4:38 PM, Mati Vapper <mati.vapper_at_gmail.com> wrote:

> Hi,
> I have an interesting issue in 3 node cluster where during the backup the
> normal processing is a bit disrupted. I didn't find much information from
> mos either from internet, so maybe here somebody has experienced something
> similar:
>
> Active%  | SQL_ID          | EVENT                           | WAIT_CLASS
>    | SERVICE_NRNAME    | PROGRAM
>
> ------------------------------------------------------------------------------------------------------------------------------
>     95%  |                 | db file parallel write          | System I/O
>    | SYS$BACKG         | oracle_at_linuxsrv1 (DBW1)
>     81%  |                 | log file parallel write         | System I/O
>    | SYS$BACKG         | oracle_at_linuxsrv1 (LGWR)
>     57%  |                 | db file parallel write          | System I/O
>    | SYS$BACKG         | oracle_at_linuxsrv1 (DBW0)
>     24%  | b1zbvg2qpwt05   | reliable message                | Other
>     | service1          | JDBC Thin Client
>     24%  | b1zbvg2qpwt05   | enq: ZZ - update hash tables    | Other
>     | service1          | JDBC Thin Client
>     19%  | b1zbvg2qpwt05   | enq: ZZ - update hash tables    | Other
>     | service1          | JDBC Thin Client
>     19%  | b1zbvg2qpwt05   | enq: ZZ - update hash tables    | Other
>     | service1          | JDBC Thin Client
>     14%  | fzbbuz7q4qpka   | db file sequential read         | User I/O
>    | service1          | JDBC Thin Client
>     14%  |                 | log file sync                   | Commit
>    | service1          | JDBC Thin Client
>     14%  |                 | log file sync                   | Commit
>    | service1          | JDBC Thin Client
>
>
> The service1 is an application service which is running on 2 instances. The
> enq: ZZ events are probably related to global context, because the sql is
> the execution of dbms_session.clear_context procedure.
>
> Thanks,
> Mati.
>
>
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>


--
http://www.freelists.org/webpage/oracle-l
Received on Fri Oct 11 2013 - 15:53:54 CEST

Original text of this message