JOB Monitoring

Hi All,
I am involved in job monitoring Activities for last few months and I am really unable to understand the mechanism.Please help me out in understanding the thing.I will assign points duly.
1)I want to know the use of jobs.Why do we require them.<b>Also what is the relation between jobs and events and how they affect the Performance of day to day loads.</b>
See,I am asking this because,sometimes what happens is a job fails, but the loads are successful.Recently ,I did not check out the jobs,But instead went into RSMO in the Production Server and found out the loads have been successful.So,I thought there wasn't any problem.But a job has failed in SM37,which I found out on the next day and I was responsible for not informing the BW onsite person about this cancelled job .So,Can anybody help me to understand the things in the above context.
I assure you,points will be assigned.It is very urgent.
Also,I will post the another scenario further in a new thread regarding the same issue,once this question is answered
Regards,
Poonam

Dear Poonam,
For every single task in BW a program is needed. Such a program can run in a so-called dialog-mode or in batch-mode. These words are quite old, they come from the old mainframe times.
In dialog-mode a program is able to receive additional information via keyboard entries and can output messages at the screen. In batch-mode both options are not available. Instead, parameters have to be given via a variant and output data is written to a file - for instance a job log.
A job consists of one of more steps. Every step represents a certain program and its variant.
According to the explanations above, something must be wrong in your situation. It is not possible, that everything went fine although a job failed. Under normal circumstances you should find additional information in transaction ST22 or SM21.
Kind regards,
Stefan

Similar Messages

Unified Job Monitoring -Share your experiences with SAP Solution Manager 7.1

Dear Forum members,
I woudl like to introduce myself as Raghav, S. I am in the development team responsible for building the unified job monitoring workcenter starting 7.1 SP10 (onwards).
Please post your feedback, current status wrt to setup, commmon issues in this area.
I will be more than glad to help you here.
Regards, Raghav
Development Manager
CLM - Solution manager team, Labs India.

Hello Raghev,
we implemented unfied job monitoring on a larger scale by now, starting out with the monitoring of SAP Standard & Reorg Jobs.
Everything seemed quiet OK at the beginning, but over time display performance of the Monitoring UI decreased drastically (exceeding 5 Min. "per click" in the UI).
I pinned this down to the following (Oracle) SQL access:
SELECT
"CLIENT", "CONTEXT_ID", "TYPE_ID", "HASH_METRIC_PATH", "START_TIMESTAMP", "IS_CURRENT", "END_TIMESTAMP",
"METRIC_PATH", "RATING_AGGREGATE", "VALUE_MIN", "VALUE_MAX", "VALUE_SUM", "VALUE_COUNT", "VALUE_LAST",
"VALUE_DATA_TYPE", "VALUE_MAX_TIME", "LAST_TEXT"
FROM
"MES_DB_AGGREGATE"
WHERE
"CLIENT"=:A0 AND "CONTEXT_ID"=:A1 AND "TYPE_ID"=:A2 AND "IS_CURRENT"=:A3 OR "CLIENT"=:A4 AND
"CONTEXT_ID"=:A5 AND "TYPE_ID"=:A6 AND "IS_CURRENT"=:A7 OR "CLIENT"=:A8 AND "CONTEXT_ID"=:A9 AND
"TYPE_ID"=:A10 AND "IS_CURRENT"=:A11 OR "CLIENT"=:A12 AND "CONTEXT_ID"=:A13 AND "TYPE_ID"=:A14 AND
"IS_CURRENT"=:A15 OR "CLIENT"=:A16 AND "CONTEXT_ID"=:A17 AND "TYPE_ID"=:A18 AND "IS_CURRENT"=:A19
The execution plan used for this SQL is:
0
SELECT STATEMENT
1
INLIST ITERATOR
2
TABLE ACCESS BY INDEX ROWID MES_DB_AGGREGATE
3
INDEX RANGE SCAN MES_DB_AGGREGATECU
MES_DB_AGGREGATE in our case has over 5 million entries.
The parameterization of the DB incl. Memory and all should be OK.
The index range scan simply takes that long.
This might be, because the MES_DB_AGGREGATECU index has the lowest cardinality column (CLIENT) first.
Header 1
Header 2
COLUMN_NAME
NUM_DISTINCT
CLIENT
1
CONTEXT_ID
433
TYPE_ID
2192
IS_CURRENT
2
Could you have your team check on this, because with the situation as is, the Job Monitoring application seemingly becomes non-usable, as soon, as you start to scale up.
Best Regards,
Thorsten

BizTalk Zombie Message not coming as error in SQL Job "Monitor BizTalk Server"

Hello All,
I have Zombie message in Message box
But when I run job "Monitor BizTalk Server" no error reported.
My understanding this should be reported as error in second step. Same has been in details in URL
http://biztalkmessages.vansplunteren.net/2010/10/19/monitor-biztalk-server-job-in-biztalk-2010/
Can anyone help why above behavior is different ?
Tarun

Well....here's the deal. It seems the group knowledge on the Monitor BizTalk Job is pretty limited. Here's a better blog with some more information:
http://social.technet.microsoft.com/wiki/contents/articles/19700.biztalk-server-monitor-biztalk-server-job.aspx
I don't think a Zombie would immediatly qualify for any of those cases because for some period of time, it still belongs to a service instance, just one that can no longer Consume it. Eventually it may qualify for condition 4 once the Completed Orchestration
is purged.
Disclaimer: This is educated speculation. If anyone has done any testing around this, I and others would appriciate some guidence.

Business Process Monitoring (Job Monitoring)

Hi all,
I have some business processes being monitoring. Most of them are background jobs and we are monitoring delays, cancellation, duration...
My problem is that we have an external tool (CONTROLM) for job scheduling.
That means that jobs are not released directly in SAP, but they are in status scheduled and when the start condition (in the external tool) is reached, controlm makes a copy of the scheduled job and start it in SAP.
I have succeded to have the alerts working checking the option of "start job condition" , but some times I receive alerts like this one: "Job Name with schedule date 16.09.2010; not yet started" even if job is started and finished correctly.
Anyone knows what could be the problem? Or why are we receiving these "fake alerts"?
Best regards,
Virginia

Hi Virginia,
if you work with external schedulers and use start condition "Start by time" then you have to ensure that the maintained Planned Start Time in BPMon is earlier than the actual Start Time coming from the scheduler. If you use start condition "Use job start condition" this problem cannot occur as every job is picked up that fits the naming criteria. But there you cannot check for the start delay (business wise) but only system wise.
In general you have to ensure that either the latest SP for ST-PI plug-in is installed on your backend system or the latest SAP notes (component SV-SMG-MON-BPM) related to ST-PI are implemented. Only then you can ensure that no "funny" behaviour occurs when it comes to job monitoring.
Best Regards
Volker

Background Job Monitoring not getting activated

Hi Gurus,
I have been trying to use the batch job monitoring with ccms.
I have a problem in activating the "Background Job Monitoring" in few systems.
I followed the steps in SAP Note 553953 but no use.
In RZ21 i am not able to see the Background Job Monitoring under "Background" MTE .
The same is wokring in Dev but in QA and Prod its not .
Please suggest the solution and cause of the problem.
Dev

Hi,
in SAP Solution Manager there are actually tools to support you in implementing the Job Scheduling Management (JSM) standard and these tools cover the request, documentation, schedulign and monitoring of a job.
http://service.sap.com/~form/sapnet?_SHORTKEY=01200252310000087418&_SCENARIO=01100035870000000202&_OBJECT=002007974700000007302009E
Visit the following FAQs to find out more about JSM
http://wiki.sdn.sap.com/wiki/display/SM/FAQJobScheduling+Management
http://wiki.sdn.sap.com/wiki/display/SM/FAQBusinessProcess+Monitoring
Kind regards,
Martin
http://service.sap.com/jsm
http://service.sap.com/bpm

How can I edit the Alert Description for a SQL Agent Job monitor, specifically Job Duration.

I'm setting up some sql agent job monitoring, I need to monitor for Job Duration.
This is easy enough, I just create the overrides and set the parameters and it alerts as expected, however there is no alert text. The alert does me no good with out an alert description. How can I change this?
Paul Arbogast

Paul
It is not an easy way for your just copy that monitor and modify it. If you has interest, you may use MP2XMLDumper tool to export SQL server monitoring MP , modify relate moniotr and import it into SCOM.
You can download MP2XMLDumper tool from
http://blogs.technet.com/b/cliveeastwood/archive/2007/08/22/mp2xmldumper-version-2-0-export-opsmgr-mp-s-to-xml-using-mp-name-export-criteria.aspx
Roger

CCMS Background Job Monitoring

We have setup CCMS background job monitoring and use it to send alerts for failed CH_* jobs. We activated an autoreaction for Job_Monitoring_Status.
Our autoreaction method has some extra functions to suppress duplicate alerts. It uses the alert text to check this. The standard alert text for Job_Monitoring_Status includes the job id + the end time: e.g.:
u201CCH_APAC_SD_RVV50R10C_DEL_BS_S1 | 03282700 | Termin, started at 12.10.2011,03:28:27 terminated at 12.10.2011,03:28:29u201D
This will let us duplicate suppression fail. We only need the job name in the alert text.
Can we change the message class for the Job_Monitoring_Status? Do we need to write our own message class?

Hi,
This is possible and perhaps the link in the thread below should help:
Simple question for CCMS Output format
Regards,
Srikishan

Anybody using Batch Job Monitoring?

Hi Team,
Looking for some Solman BPM Batch Job monitoring help with the "job not started" monitor. It's not detecting the alerts.
Thanks,
Milton

Dear Milton,
For setup and monitoring possibilities, see the Business Process Monitoring Set-up and User Guideline Business Process Monitoring with SAP Solution Manager Releases 7.1 SP05+,
as of page 65, e.g. chapter 6.1.2.1.2 Start procedures in detail.
Kind regards, Barbara

Applying Business View format of ARA into Job Monitoring of Process Control

Hi Gurus,
I have an issue regarding with the Job Result of when using SOD sub-scenario in Process Control. We have integrated AC and PC and the controls that have been set for automation, however, the fields (User ID, Name, Group, Risk Description, etc.) are not showing on the PC Job result like the ARA can do. Below information are given:
1. In ARA --> User Level --> Simulate the risk analysis with Business View format --> Result --> choosing Management Summary --> above mentioned fields are shown.
2. In Automated Monitoring --> Choosing Management Summary --> onlyfields of technical view format are the only ones that can be shown (User ID,Risk ID and Control ID).
Thus with this, is there any possibility that the Result of the Job Monitoring of the PC can be manipulated? If yes, can you provide some steps?
Hope you can help me on this.
Kindly refer on the attached file for more info.
Thanks!
Mackoy

Christopher,
Re-generating the transfer structure does not need InfoPacks to be re-created. if you are still getting the old error, two possibilities:
1) You created a new DataSource with new transfer structure, etc. but the InfoPack in the process chain still points to the old trsnfer structure.
2) The data in the table has some records as MM/YYYY and some as YYYYMM.
You don't need to re-create the InfoPacks if your transfer structure is re-generated. They should work off the latest definition always.
Cheers
Aneesh

Issue in periodic jobs monitoring

Hi Guys ,
i have created a program for monitoring periodic jobs. out of list of some 120 predefined periodic jobs if the job is not released (Status S in table TBTCO) then send a warning mail with details of the jobs.
this program is set as a daily job. now problem is that when my program runs through job then it sends some jobs in mail though they r released...while manually checking in table TBTCO i found 92 jobs are in released status but if checking after 5 mins its showing 91 jobs in released status....can any one know why these periodic jobs status behaves randomely... is it like if periodic job is delayed then its status is not released...or do any one have any idea regarding this abnormal behavior???
Kind Regards,
Jacks M.

My Dear Experts,
It is my select query hitting at a particular job changing it's status for released to Ready/Active in which the difference might be in Micro/Nano Seconds this has caused the job monitor tool output showing the relevant job as Not Released... Thanks guys for being patient enough to a query which might be insufficient in getting ur minds work...
Kind Regards ,
jacks.

Data Services Admin Console "Read Timed Out" via Job Monitor Log Tab.

Hi All,
I am getting the "Read Timed Out" via Job Monitor Log Tab issue, when large ETL jobs are being executed.
Is there a resolution for this issue.
Thank you
Kind regards
Hai.

Hi Manoj,
I have checked the UNIX server and the admin.xml is not located in advised location.
The only two file that are in the advised location are:
- ContentTypes.xml
- PythonAPI.xml
I have another install of data service (default installation) on a windows environment, and I checked the advised location and the file is there.
Is it possible to copy the windows admin.xml to the Unix server.
Thanks
Kind regards
Hai.

Unified Job Monitoring: How to delete a Monitored Object ?

Hello,
in unified Job Monitoring I face the Problem, that I deleted a Monitoring Scenario without deleting the included Monitored objects.
I set up another Scenario and would like to reuse the Monitord object names - which is not possible, because the System informs me, that they already exist. Also, in the Monitoring UI These Monitored objects still Show up and there's no way to get rid of them.
Thanks for your Support !
Best Regards,
Thorsten

Do you mean bc4j.xcfg file? It will be under your app. module ./common sub-directory.

Extra Logon required when accessing the Job Monitor

Hi Friends,
This is an annoying problem. Whenever I load data to a data target ie Infocube or ODS, upon accessing the job monitor from the Manage target screen, I am prompted to present login credentials even though I am already logged in.
My account is configured with SAP_ALL authorizations.
Does anyone know how to stop this intermediary logon step ?
Thanks
Nate

Information from the HTTP access log:
143.97.201.128 - - [28/Aug/2006:10:22:13 +0200] "GET /SAP/BW/DOC/META/FLDMETA/MIS_HSE_DD_HELP2.ppt HTTP/1.1" 200 627200
143.97.201.128 - - [28/Aug/2006:10:22:16 +0200] "OPTIONS /SAP/BW/DOC/META/FLDMETA/ HTTP/1.1" 401 2042
As you can see the first GET to the powerpoint file gives a return code of 200, everything OK.
Then a cpl of seconds later there is a "OPTIONS" call that returns 401, access denied, which produces the logon dialog box in Internet explorer.
Help!

CCMS Background Job Monitoring Runtime Alert

Our Basis team has configured CCMS and the Background Job Monitoring for long-running jobs. They want to set up a "generic" entry (for example, * for job name) so that all jobs will be monitored for runtime longer than 1 hours (will be longer duration when moved to production). This seems to be working correctly but the alert message does not give the name of the background job that triggered the alert. The altert text is "Alert Text:3821 Sec > 3600 Sec Current value over threshold value System:DEV ". This message is coming from message class RT, number 001. How do we get the job name included in the alter message? I have looked at the messages available in message class RT but do not see one that includes a variable for job name. We do not know where to look to see what data is actually available for the alert text message and what other message classes might be possible.

Hi,
This is possible and perhaps the link in the thread below should help:
Simple question for CCMS Output format
Regards,
Srikishan

Background job monitoring - alert from second failure

Hi All,
There is a need for configure monitoring for background jobs.
I have been enabled background job monitoring via SE16 and background job is visible in RZ20.
The problem is that we want to have a critical alert from second failure of certain job, not from first failure.
Seems that I need to create a new method and assigne it to background job in RZ20?
Is there a documentation how to create own method? or is there a another way to implement this monitor?
Best Regards,
Jani Mäki

Hi,
Do you want to monitor second failure on a day?
Do you know the schedule of second job run?
If yes to these questions, you can try to monitor using BPM in solution manager.
Feel Free to revert back.
-=-Ragu

How to clean /Delete all the jobs from Job monitor.

Hi,
we have done lot of testing in redwood. we have scheduled many jobs through redwood.we want to clean up the Job monitor now.we want to delete all the jobs from job monitor.
1)Is there any standard system give Job definition to do so?
2) how to delete jobs in waiting status?(next run is going to lock wait due to this waiting job status)
3)how to delete all the Operator messages?
Regards
Rajesh

Hi,
Yes that was for individual job definitions but often if there are many jobs, most jobs come from only a few job definitions (SAP_AbapRun, SAP_CCMSJob, ...) so these options could help you already to delete most jobs.
You could also create a filter showing all Waiting jobs, then select all jobs, right-click and kill all these jobs.
Then a filter for all jobs in an end-state (Completed, Error, Killed, Canceled), select all jobs, right-click and delete them.
Regards,
Anton.

JOB Monitoring

Similar Messages

Maybe you are looking for