CCMS job delay monitoring not working

[Former Member]

Hi folks,

my goal is to setup job delay monitoring via CCMS, I've done everything according regular guide e.g.http://help.sap.com/saphelp_nw2004s/helpdata/en/1c/48803d48de0610e10000000a114084/content.htm

Plus I've also modified properties of "Delay" to be 3 seconds for critical

However it does not report any problem - status is green

even when job has delay in fact (recently 6,7 seconds, but ussually significantly more...30-50sec)

Have not found any clue, everywhere is written, how easy is to set it up 😕

What could be wrong? Something here?

Thanks in advance for your comments.

BR Miro

[Former Member]

Hi,

I am not exactly sure if you want this  but please try this.

RZ20> SAP CCMS Admin Workplace-> Background Processing (Double click)

And then check the internal nodes.

BR,

Prabhakar

[Former Member]

Dear

I believe u r looking for delay in the background jobs.If this is correct ,why would there be delay in active job.The job changes to active only when it gets free background work process.Before that the status is released with delay.So u should monitor the no of free background work process or any other relevant monitor. 

[Former Member]

Hi Prabhakar,

I can't find it, there's no

RZ20 > Background Processing (Double click)

nor

RZ20 > SAP CCMS Admin Workplace > Background Processing (Double click)

If possible, please describe exact guide.

Thanks

[Former Member]

Hi Chander,

that's correct, I'm looking for delay in the background jobs. For test purpose I've picked up SAP_CHECK_ACTIVE_JOBS background job, which runs every hour. I think it does not matter what the job exactly does, but at the moment I want to know if this job was started with delay or not. FYI I tried also other periodical job - nothing has changed, so the problem is not with job "SAP_CHECK_ACTIVE_JOBS" but probably somewhere in settings or so...

[Former Member]

The active jobs will not show show any delay.Try with some other parameter say file system and then check

[Former Member]

Hi Chander,

sorry but I don't get what you mean. Why it shouldn't work, when there is such feature? SAP_CHECK_ACTIVE_JOBS must behave same (start/stop/after event/duration/delay) as any other job in system, no matter what exactly they does.

It's proven already, that it is working - because I can see it in alert history.

When I listed job history, we can see there was "jump" between run at 22:00 and at 23:00

So it brings me back to "Delay" properties -> tab PerformanceAttributes and section "Comparisn value" -> which is at the moment set on "Last reported value"

However when you compare 189 sec. delay at 22:00 and 46 sec. at 23:00 the difference isn't 91 sec. as reported above. It suppose to be 143 sec., so where are those 52 seconds? That would also explain, why small differencies are not reported at all, even when threshold values are set to 1+3 seconds.

This couldn't be normal, right? Is here someone, who has similar scenario working properly? At the moment it proves, that it's working but not as fast as I want (3 seconds for critical alert). I can live without that, as I will not need such strict values, anyway I'm still curious about that.

[IanSegobio]

Hello Miroslav!

I would like you check first in the transaction RZ20 if for the monitor segment in which the background monitor context is located, enough slot free space is available for new alerts, as described in note 731165:

First, find out what is the correct monitoring segment for background:

---> RZ21

     ---> (Topology) Context overview

Scroll down until you find "Background" in the first column. Then, in the same line (third column) you'll find the correct segment to check later, in transaction RZ20.

Second, checking whether there is enough free space for new alerts:

---> RZ20

     ---> SAP CCMS Technical Expert Monitors

          ---> System / All Monitoring Segments / All Monitoring Contexts

               ---> <Instance_name>

                    ---> MoniInfra_<segment_name>

                         ---> Space

                              ---> Click twice in "FreeAlertSlots".

Here, you'll be able to see if there are free space available for news alerts to be created:

-------------------------------------------------------------------------------------------

Alerts                         1608 (  8997 )     ( 82.1 % Available  )

-------------------------------------------------------------------------------------------

I hope this helps!!

Cheers!!

[Former Member]

Hi Ian,

thanks for your reply. Our Alerts values are 7647 (8996)   (15% available). There is also alert reported in history with "action_required" status under "AlertsFrequency".

15.10.2013    14:53:26    <SID>    MoniInfra_hostname_<SID>_SysNr    Space    AlertsFrequency    ACTION_REQUIRED    Frequency 1017,380000 /min > 100,000000 /min  (15 min. avg. value above threshold value)

So is this one the problem? How to clean it up?

Additionally I'm attaching screenshot from current status (witnout any action at the moment)

[IanSegobio]

Hello Again,

If there are still free alert slots available, I think this should not be the issue. Could you provide a new RZ20 screenshot, with the exact same structure expanded for SAP_CHECK_ACTIVE_JOBS but this time instead of "current view" (which you've used on the already provided hardcopy) use "status data collector" view.

This can be set through menu "Views" >> "Status Data Collector"

Cheers!

[Former Member]

Hi Ian,

again, looks fine I guess.

[IanSegobio]

Hi Miroslav,

Are these MTE being updated correctly as per the job's executions? If you select the "delay" MTE and click as per below image, can you match the past delay information from RZ20 with the same information provided by SM37?

Cheers!

[Former Member]

Hi Ian,

it seems it does check every 300 seconds (12x per hour) so this looks correctly configured, but still doesn't catch any "delay"

SAP_CHECK_ACTIVE_JOBS runs once per every hour.