CDMC Performance and Memory related query

[Former Member]

Hi Experts,

I am looking some guidance on CDMC functionality of sm 7.0 ehp1.

as per the guide, it was no where mention how long time it is required to finish the CA and UCIA? what are the constraints would be considered here?

2. If I scheudule 3 months statistics data collection from production, does it impact my production system performance, hw can I assure that no performance issue in production due to CDMC satistics collection?

3. If my collected data stored in CDMC database table in solution manager then, do I need to look the free space in solution manager, if so after the analysis how can I clear the occupied memory? any house keeping jobs?

Please share your input based on your experience.

Thanks,

Jansi

[bjoern_panter · Accepted solution]

Hi Jansi,

@1) the runtime depends on the overall number of custom code objects, the available hardware and the executed analysis steps. OIn most cases the results are availbale on a very detailed level with in 2-3 h.

Sometimes it needs 20h if the hardware is weak.

@2) the is NO impact on production. Thats the biggest benefit. We collect more than 3 month. Just activate the collection now. Also if the ST03N data is using only 3 months, CDMC will collect it for a longer period.

@3) doesn't matter about the DB soizes. It is very small, cause SAP is using latates available compression algorithms to achieve best runtime and best space consumtion. We save the results in analysis projects. These can be deleted without any issues.

regards

Björn

[Former Member · Accepted solution]

Hi Bjoern,

Thanks, Its clear now.

regarding Q2. can you explain more on this.

How long solman takes to get the 3 months data from Production, As per my understanding it is the another extraction job from solman to prodution,

Bjoern Panter wrote:
 
 
 Also if the ST03N data is using only 3 months, CDMC will collect it for a longer period.

I was not able to understand this, please can you clarify.

Thanks,

Jansi

[Former Member · Accepted solution]

I think it has to do with the fact that CDMC does not technically rely on the ST03N data to pull information, rather than collector jobs that run independently.

I may be wrong, but just what I understood.

An absolute scorcher of a topic mind you! Great work!

Cheers Bjoern for bringing some much needed focus on these tools.

Trust the makers of ABAP indeed

[Former Member · Accepted solution]

Hi,

yes, this is what Bjoern trying to tell it seems.

now again, if this case, st03n has 1 month data, whereas CDMC can collect 6 months data? if yes from where it collects? I thought ST03N is the data source for CDMC.

I guess still I need better explanation for understanding.

Please help me to integrate.

Thanks,

Jansi

[juergen_mahler · Accepted solution]

Hi,

that's quite simple:

CDMC starts 3 periodic jobs to collect usage information: CNV_CDMC_CA_COLLECT_STAT_700  ==> ST03N , CNV_CDMC_CA_EVAL_SHORTDUMP ==> Short dumps  and  CNV_CDMC_CA_COLLECT_TAB_STAT==> ST10.

The collected data are saved in own CDMC tables in the managed system, that means CDMC can collect everything available , e.g. from ST03, saved the info in own tables and recollect periodically the delta based on your settings in the statistic collection of CDMC.

Does that help?

Regards,

Jürgen 

[Former Member · Accepted solution]

HI Juergen,

Let me be a bit more of a noob and extend on this statement a bit.

Does this mean if I enable the jobs now in July 2012, I will be able to look up data from Jan 2012 or earlier , or only from July 2012.

I think SJ is trying to understand if there is any way to access previously collected information or if indeed it automatically collects from the day of "enablement".

Cheers!

[juergen_mahler · Accepted solution]

Hi,

Depends on the system settings for ST03n. CDMC is able to get all available months specified in st03n and is doing the next month period by period...

Example: ST03n: available months 04 / 2012 – 06 /2012

1. Start CDMC 07/2012                   get the data 04 / 2012 – 06 / 2012
2. Start CDMC 08/2012                   get the data 07 / 2012
3. Start CDMC 09/2012                   get the data 08 / 2012

Regards, Jürgen

[bjoern_panter · Accepted solution]

If you are able to re-generate a usage statistic for the last 6 months, then I guess you will be the sorcerer of SAP. 🙂

Default usage data collection is periodic 3 months. This is more or less fixed. And housekeeping will delete old data every months with the FIFO principle.

CDMS itself avoids this. As soon as you start the statistical collection, the system will read every month the latest available data, compress it to custom code aspects and transport the delta to the solution manager system.

Only with this method you will have the possibility to analyse more than 3 month default. Here you will have the real benefit compared to Panaya. They work with only 3 default months.

We can work with an unlimited number of monthly data, as long as you have activated it early enough.

So in best case activate now and wait some time. The default 3 month can be evaluated within the next 15 minutes. No setup needed, just click the activties in CDMC.

Next month you will have the data for 4 months and the months after you data collection will have 5 months.

and you mentioned it already: Trust the makers of ABAP 🙂

If you have any improvement ideas, we can implement it.

[Former Member · Accepted solution]

Hi Bjoern,

This is what I actually looking, Now I am clear, giving this update to my clients,

Once again Thanks for the time taken and updated.

Thanks,

Jansi