Ansicht
Dokumentation

12547 - MTXW Status and Job QGYSERVER

12547 - MTXW Status and Job QGYSERVER

ABAP Short Reference   ABAP Short Reference  
This documentation is copyright by SAP AG.
SAP E-Book

MTXW Status and Job QGYSERVER

Hello Volker,

I found some SQL 913 errors on table D010SINF in ST22 from workflow-
related program and I will open an OSS message later.

I forgot to mention the timing of the MTXW statues. It last from one
to five seconds most of the times (based on the experience of my
clicking F10 key when WRKACTJOB SBS(R3_xx)). Sometimes it could be
as long as 10 seconds or more for some work processes (not all of
them). We saw all the dialog work processes are running in Tcode
SM50 a few times last week and suspected that it was related to some
kind of locking issue. But so far we have not found directly link of
it to MATX status.

After using a F5 macro, I caught the following objects in mutexes
monitor:

DBMON_HDR1
DBMON_DATA01
DBMON_DATA04
DB_TBUFF

They look related to DB monitor. (It might be normal because they
didn't stay long when the F5 macro was running.)

Can we verify that our DB monitor (also performance monitor?) is
configed correctly to give enough info to ST04 and have minimal
overhead? I checked the redbook again and didn't find the suggested
config of it.

In the business hours, we have CPU usage 50% - 80%, DB capability 15-
20%, 0 for Wait-Inel and Act-Inel, about 50 DB faults and 300 Non-DB
faults in 19G *BASE pool from WRKSYSSTS and disks about 5%-8% busy
from WRKDSKSTS. Do you think we are close to limit of the hardware?
(Based on a false assumtion that we are updated to everything and
have fine tuned the ABAP programs.) If yes, should we say that CPU
is the bottleneck?

Thank you very much,
Victor


--- In DoNotReply@consolut.eu, "Volker Gueldenpfennig"
<volker.gueldenpfennigZs...> wrote:
> Hi Victor,
>
> MTXW has to be analyzed for the root cause. Try WRKACTJOB 5 and
then
> 19. Then you should make a hardcopy and open an OSS ticket with
that.
> If it is QQQDBMTR, it is caused because of the DB-Mon. Otherwise
it
> will be caused because of SAP somehow. We would have to
investigate.
> QGYSERVER really looks like SQL0913 ...
>
> But, the SQL0913 could be caused because of the long MTXWs in
other
> WPs. Often this is happening together.
>
> Regards
>
> Volker
>
> --- In DoNotReply@consolut.eu, "Lin, Victor" <vlinZp...>
wrote:
> > Hello everyone,
> >
> > After we upgraded OS/400 from V4R5 to V5R2 with CumPack 3252, DB
> Fix 10, and other individual PTFs and upgraded SAP kernel to 1617,
> we noticed that the dialog work processes are in MTXW status more
> frequently than ever. From F1 help (The initial thread of the job
is
> in a mutex wait. A mutex is a synchronization function that is
used
> to allow multiple jobs or processes to serialize their access to
> shared data.) it looks like a locking wait. But why?
> >
> > Also we are getting jobs QGYSERVER under the subsystem R3_XX. We
> used to get those jobs when we are applying the support packages
> with user activities on the system, and they went away days after
> the support packages were finished. But now (10 days after the
last
> time we applied the support packages) there are still about 20
> QGYSERVER jobs in DEQW status which entered the system about 1 or
2
> hours after midnight in the last a few days.
> >
> > Meanwhile we found and fixed a few slow transactions like MB52.
> But more transactions, for instance V/LD, VK34, and MM17, shows
> performance issues. Does anyone have similar experiences with them?
> >
> > Any help or hint is appreciated!
> >
> > Thank you very much,
> > Victor
> > SAP Basis Manager
> > Plastipak Packaging INC


Durban Tours - Südafrika Safari

SUBST_MERGE_LIST - merge external lists to one complete list with #if... logic for R3up   ABAP Short Reference  
This documentation is copyright by SAP AG.

Length: 4620 Date: 20240329 Time: 152744     sap01-206 ( 4 ms )