DISP+WORK Shutdown Immediately

[former_member184114]

Dear All,

I am facing an issue from past three days on our production server. Following is the explanation for the same:

We daily take backup from OS level in the night and start our SAP server after that(this is done through windows scripts, which are scheduled tasks). This has been a practice for a very long time with out any problems. However, last thursday, after backup when SAP started, I could see that disp+work process got stopped.

Upon further investigation in disp+work log file in /work directory, I could see the below details:

---------------------------------------------------
trc file: "dev_disp", trc level: 1, release: "46D"
---------------------------------------------------

Thu Jun 11 08:47:55 2009
relno      4640
patchlevel 0
patchno    2337
intno      0
pid        3404

***LOG Q00=> DpSapEnvInit, DPStart (00 3404) [dpxxdisp.c   921]

Thu Jun 11 08:48:01 2009
*** WARNING => DpNetCheck: NiAddrToHost(1.0.0.0) took 6 seconds
***LOG GZZ=> 1 possible network problems detected - check tracefile and adjust your DNS settings [dpxxtool2.c  3171]
MtxInit: -2 0 0
DpIPCInit: start server >syeungrp_GRP_00     <
MBUF state OFF
MBUF opmode USE
EmInit: MmSetImplementation( 2 ).
<ES> client 0 initializing ....
<ES> InitFreeList
<ES> block size is 1024 kByte.
Using implementation std
<ES> Memory Reset enabled as NT default
<EsNT> EsIUnamFileMapInit: Initialize the memory 2500 MB
<ES> 2499 blocks reserved for free list.
ES initialized.

Thu Jun 11 08:48:02 2009
***LOG Q0K=> DpMsAttach, mscon ( syeungrp) [dpxxdisp.c   8384]
MBUF set hwid_state to HWID_PENDING
DpStartStopMsg: send start message (myname is >syeungrp_GRP_00     <)
DpStartStopMsg: start msg sent
CCMS: AlInitGlobals : alert/use_sema_lock = TRUE.
CCMS: Initalizing shared memory of size 10000000 for monitoring segment.

Thu Jun 11 08:48:03 2009
CCMS: start to initalize 3.X shared alert area (first segment).
DpMsgAdmin: Set release to 4640, patchlevel 0
MBUF state PREPARED
MBUF component UP
MBUF set hwid_state to SAP_O_K (Y0354306574                             )
DpMsgAdmin: Set patchno for this platform to 2337
Release check o.K.

Thu Jun 11 08:48:11 2009
MBUF state ACTIVE

From the above log, it was obvious that there was some Virtual memory issue. However, we maintain the following virtual memory settings: RAM=2GB; Virtual Memory=4GB

[reza_memari]

Dear Faisal,

Are you getting any error in your windows event logs?

[former_member184114]

Dear Reza,

We got windows error like:

Windows cannot load the user's profile but has logged you on with the default profile for the system

It seems that system is unable to load the <SID>ADM user. However, we have almost 1.5GB of free space in C:\ drive and we dont use this drive for any purpose.

In order to solve the above problem, if I restart our production server for O/S level, then it seems to be ok.

Regards,

FAisal

[Former Member]

Hi,

First of all as recomended the Application should come down first then the Database.Your stoping process of sap system told by you is wrong.

And for the user properties go Manage and the user management and check the setting of the sidadm.

Thanks

Rishi Abrol

[former_member184114]

Hi Rishi,

It is my typo error. the script shuts down Application first and then Database.

Based upon the investigation so far I have done, it seems windows is unable to accomodated <SID>ADM user profile. That is why, when we log on using this user id, we get the error I mentioned above.

This is being resolved with relevant department, I will update you here once I get the solution.

Mean while, if you have any suggestion, please let me know.

Regards,

Faisal

[former_member184114]

Hi,

We have deleted some of the unused files on primary drice (C:\) and users' profiles which were available for long. After that, I restarted the server from OS level. This solved our problem for now.

Is this the correct way and the real cause of the problem? This is still to be assertained.

Any way, for now this problem is solved.

Thanks all for your contribution.

Regards,

Faisal

[former_member184114]

Further to my above post, below is the additional information:

This is a standard settings and we did not face any problem so far with this settings until last thursday.

When I restarted Windows Server (from OS level) and restarted SAP, then it seems every thing was fine and this time disp+work process did not stop.

I have gone through the following notes:

525900, 124562 and 501902

In Note#501902, it says the corrections are included in Patch 46D:1217

If you could see above, we are at patch level 0 and patch numer=2337

May I know the difference between the above 2?

If this is the case, then my question is why we are facing problem all of a sudden? Though system was running successfully for so long. And there is no change in any component as well.

Also let me tell you, our DEV and QA systems are on the same patch levels and patch nunbers. We did not face any problem on these servers.

Now, daily I have do restart Windows Server and restart SAP to solve this problem which is not appropriate. Can anybody help me understanding this?

Your contribution in understanding and solving the above issue is appreciated.

Regards,

Faisal

[Former Member]

Hi,

Please let me know these details.

Are you getting the system offline.

Is there cleanipc command in you script.

Did you install and new patch or any activity on production.

How are you stoping the system.Are all the processes getting stoped when you stop the system.

for your Question the patch level is the number of patches that sap release.

and the Number is some patch coding that has would have kept.

Its the patch levelthat we see at the service market place and check the new release.

Thanks

Rishi Abrol

[former_member184114]

Rishi,

Thanks for your response. Following are the answers to your questions:

Are you getting the system offline.

Yes. the script what we have stops DB and then SAP.

Once the backup is over, the start script starts DB and the SAP.

Is there cleanipc command in you script.

No, there is no cleanipc command in the script

Did you install and new patch or any activity on production.

No, we did not install any new patch.

How are you stoping the system.Are all the processes getting stoped when you stop the system.

As I you earlier, these are scheduled tasks in windows. So, really I cannot tell you exactly whether all the processes are getting stopped properly.

Should you need any more informatin, please let me know.

[Former Member]

Hi Faisal,

Clearly a network problem.

please check all your network settings and also check the host file and update your system IP in it with the hostname.

i think this should resolve it.

[Former Member]

Faisal,

Check the Phys memsize or extended memory and also check the pagefile size to resolve the issue.

and also before restarting SAP next time, do the cleanipc also for the instance.

Hope that will resolve you issue.

Regards

Ravi