Solution Manager technical monitoring - PROBLEMS

[symon_braunbaer]

Dear Experts,

I am going nuts on the Technical Monitoring with SAP Solution Manager 7.1 SPS 12.

I have spent sooooo much time to configure things, and again, there are quite a few

frustrating errors.

No matter what I do, I keep receiving fake alerts about systems down, whereby the

systems are up and running.

When I click on the Managed Object column in one of those fake alerts, I am getting

the following very nice message:

Application does not exist

Message no. CL_URLAPI_MESSAGE101

Kindly advise what does it mean and how it might be solved. Many thanks in advance!

[ruth_reilly · Accepted solution]

Hi Symon,

Please review the following SAP KBA reducing the possibility of getting false alerts

2046931 - Grey ABAP instance availability metrics - most common
issues in technical monitoring with ...

Regards

Ruth

[symon_braunbaer · Accepted solution]

Hello Ruth,

sorry for my delay, but there are too many problems to handle currently 😕

I have checked the note to which you are referring, but the first couple of lines are very poorly

describing how to get to those metrics, so I wasn't able to find where to go.
I tried transaction sm_workcenter, then selected Technical Monitoring, but there everything is just text. there are no "lamps 🙂 " which would show grey or green or any color. Kindly guide me on where to check this.

Thanks a lot!

[symon_braunbaer · Accepted solution]

Hello Ruth,

my previous post is still valid, I am still unable to navigate to the area described in the note.

Nevertheless, I found that there is a problem connecting to sapstartsrv. So I went detailed and systematically through this point of the note:

For the case Cannot connect to sapstartsrv, the note says:

Cannot connect to sapstartsrv:

Note: 1890700 - System Monitoring UI shows 'Cannot connect to service sapstartsrv' error

Note: 1757091 - Incorrect Role Assignments for MAI (-SP8)

Note: 1771916 - Metrics collected via SAPStartSrv are grey/wrong → this note is valid also for higher SP’s , perform the workaround described.

So, in detail:

Note 1890700 says:

1. Make sure the corresponding SMD Agent is up and running -> yes, it is! On both DB and AS host.

2. Verify if the SAP Start Service (sapstartsrv) is running on the managed host and if an http listener is active for port 50013 -> yes, it is! I got to the Window Alert Details: "Java instance not available" at "<SID>~JAVA~hostname_<SID>_01". There I have expanded the Alert Description and Analysis area and I have clicked View instance status in SAP Management Console. The SAP MMC console opens and shows everything green!

3. Implement the recommendations from SAP notes 1757091 and 1771916 -> 1757091 is obviously for the Solution Manager, as it says that the ST component is affected. The note cannot be implemented in our Solman 😞

1771916 says that it is about LM-SERVICE 7.10, and we have a managed system v. 7.31, so no LM-SERVICE is present... Or perhaps it is meant to be done on the Solution Manager ?!?!?!? I those notes they NEVER say where to apply the note...

4. Reconfigure Technical Monitoring for the affected system -> I have already tried that many times, it did not help at all...

The only thing that I might try, would be the following point from Note  1771916:

1. add the following line to the instance profile (or default profile)

"service/protectedwebmethods = SDEFAULT -J2EEGetProcessList -ReadLogFile -ABAPReadSysLog -ListLogFiles"

This allows to access the functions J2EEGetProcessList, ReadLogFile, ABAPReadSyslog and ListLogFiles to be used without authentication.

Here I have 2 concerns:

1) Will it work for J2EE 7.31, when the Note is about v. 7.10 ??

2) It is not a good idea due to security, but if it REALLY works and if it is the only way, then we will go for it.

Please kindly advise, I am really really lost here...

[ruth_reilly · Accepted solution]

Hi Symon,

Sorry for the delay in getting back to you.

For this Step:

In case you are getting only sporadic alerts (red alert which turns green after

fiew minutes), you can change the Threshold of

the affected metric from “Already Rated” to

“Already Rated (best rating of blast n
measurements)”. In this case single sporadic failures would not generate

alerts and this would reduce the possibility of getting false alerts.

Please go to Solman_Setup->Technical Monitoring->System Monitoring-> Step 4 Template Maintenance

Edit Mode:

For a Custom Template, on the Metrics Tab, select a metric-> you can select change settings and in the Threshold Tab below, you can change Default Threshold Settings here.

Please also see below for your reference:

http://wiki.scn.sap.com/wiki/download/attachments/269157890/Setup%20System%20Monitoring%20and%20Aler...

Regards,

Ruth

[Former Member · Accepted solution]

Hi Symon,

Please re-check the Data Collection Method and defined Expressions in Metrics Creation.

I have faced the problem and maintained correct data then problem has resolved.

Thank You,

Nahid.

[symon_braunbaer · Accepted solution]

Hello Ruth,

many thanks for your detailed reply. It is more or less clear what I need to do with the templates, just please kindly lets discuss the "sporadic" term in more detail. Do you understand under a sporadic error an error which keeps coming and going ? If yes, then how can I check, if the error is sporadic or not ? For now I can only see in the Alert Inbox the system as unavailable and a count of a few thousand occurrances, like 3680 times for example.., Also it sends from time to time an e-mail, that the system is unavailable. Thanks!

[symon_braunbaer · Accepted solution]

Hello Mohammed,

where exactly ? Could you please kindly post the transaction name and a screenshot ? Many thanks in advance!

[symon_braunbaer · Accepted solution]

Hello Ruth,

unfortunately it is not a sporadic error! No matter what I do, the technical monitoring still reports the SCS instance to be unavailable! I have set the profile parameter

"service/protectedwebmethods = SDEFAULT -J2EEGetProcessList -ReadLogFile -ABAPReadSysLog -ListLogFiles"

Also, I have tried to restart the service, as SAP support recommended in the OSS message that I have opened:

SIDadm> sapcontrol -nr 00 -function RestartService

09.12.2014 09:44:54

RestartService

OK

Then I have Confirmed the alert in the alert inbox and STILL no luck !!! The Java ICM Status still says "Cannot connect to service sapstartsrv" !!!

I am sooo frustrated and really have no idea what to do !!! PLEASE help me !!! Thanks...

[Former Member · Accepted solution]

Hi Symon,

Please check the below screen shot . We have to check the parameters in Directory browser.

Thank You,

Nahid.

[symon_braunbaer · Accepted solution]

OK, after a 3 months writing with SAP in an OSS message, I finally resolved the problem by myself! Everything they were proposing was wrong or simply didn't solve the problem.

[former_member269828 · Accepted solution]

Hi Symon,

Am facing the same issue in my ABAP system.

"Cannot connect to service sapstartsrv" alert is being generated though the service is available and running fine.

I have checked notes : 1890700 and 927637.

How did you resolve this issue?

Could you please give me the procedure that was followed to resolve the issue?

Awaiting your response.

Thanks in advance.

Best regards,

Deepika

[symon_braunbaer · Accepted solution]

Hi Deepika,

set the following parameters in your instance profile:

service/protectedwebmethods = DEFAULT

service/admin_users=<diaguser>

where <diaguser> is the user under which you are running your diagnostics agent. Usually this is daaadm. Then restart your sapstartsrv.

[former_member269828 · Accepted solution]

Hi Symon,

Thank you for the response.

I have checked the mentioned parameters and they haven't been maintained.

Should these parameters be defined in the Solution Manager system or in the respective Managed Systems?

Thanks in advance.

Best regards,

Deepika R

[symon_braunbaer · Accepted solution]

Hello Deepika and sorry for the big delay... I wasn't aware that you have additional questions.

I have always been frustrated, that on topics about the Solution Manager, the authors do not say exactly if the action has to be performed in the Solution Manager system itself, or on the satellite system. So, now I made the same mistake 🙂

So, you have to do this on the satellite system. As the diagnostics agent has to be able to connect to sapstartsrv and obtain its status, then it has to push it to the Solution Manager.

Then it takes time, till the status updates to green, so be patient and keep clicking that refresh button 😄

Hope this helps!

[Former Member]

Hi Symon,

Can you please provide us with the solution so that will be a great help to resolve simmilar issue.

Regards,

Ram

[jensfieger]

Hi Symon,

what do you mean with 'fake alert'? Usually you receive alerts like "ABAP System not available" - in this case the managed object column will route you to the webgui:

http://host:50000/sap/bc/gui/sap/its/webgui?sap-language=EN&sap-client=001

Rgds

Jens

[symon_braunbaer]

Hi Jens,

By fake alerts, I mean that the Solution Manager keeps saying about 2-3 systems, that they

are DOWN, but they are not!! I have fixed soooo many things!! The managed system

configuration, LMDB, I had wrong templates assigned (J2EE 7.00 for to a 7.31 system!) which didn't want to go away, I fixed that one as well. In the Alert Inbox, when I click the managed object, it is now opening the Web GUI, I can login, so it is obvious that the system is up, but the Solution Manager

keeps spamming with the message that the system is not available. I am sooo frustrated about that, I am now fighting 2 full days with this and still nothing...

[jensfieger]

Hi Symon,

I'd recommend to paste some screenshots of the Alert Inbox, the Alert Details and also the Overview of the System Monitoring Tree.

In one of our test systems I got an alert "Java instance not available". Indeed the managed object column references to the java start page which is available. If I check the alert in detail I can see that only one java application is affected.

Rgds

Jens