Situation monitoring for TSM does not alert when the condition is true

bananabank

Active Newcomer
Joined
Mar 25, 2013
Messages
7
Reaction score
0
Points
0
PREDATAR Control23

I create situation via TEPS GUI for monitoring TSM (kskagent) as GSMA recommend as below

I need to monitoring Schedule Status with the formular
*IF ( ( *VALUE KSK_SCHEDULE.Schedule_Status *EQ Failed *AND *VALUE KSK_SCHEDULE.Node_Type *EQ 'SERVER' ) *OR ( *VALUE KSK_SCHEDULE.Schedule_Status *EQ Missed *AND *VALUE KSK_SCHEDULE.Node_Type *EQ 'SERVER' ) )



but I found the issue situation does not trigger and alert when the condition is true

example
on my schedule workspace have information client backup failed. but not have any alert occur on my TEPS tsm instant

does some one have any suggestion to help me.

Thank you for all answer
 
PREDATAR Control23

I create situation via TEPS GUI for monitoring TSM (kskagent) as GSMA recommend as below

I need to monitoring Schedule Status with the formular
*IF ( ( *VALUE KSK_SCHEDULE.Schedule_Status *EQ Failed *AND *VALUE KSK_SCHEDULE.Node_Type *EQ 'SERVER' ) *OR ( *VALUE KSK_SCHEDULE.Schedule_Status *EQ Missed *AND *VALUE KSK_SCHEDULE.Node_Type *EQ 'SERVER' ) )


but I found the issue situation does not trigger and alert when the condition is true

example
on my schedule workspace have information client backup failed. but not have any alert occur on my TEPS tsm instant

does some one have any suggestion to help me.

Thank you for all answer


For More information

In sk log it's seem like sk agent cannot insert some data on table AGENTLOG (May be it's concern with permission of user tsm instance) as error below

2013-04-30 10:45:31.000000 TSMDerbyDB: Message: An SQL data change is not permitted for a read-only connection, user or database.
2013-04-30 10:45:31.000000 TSMDerbyDB: Error Inserting message into the Database: INSERT INTO AGENTLOG (SERVERNAME, DATETIME, ATTRIBUTEGROUP, MSG) VALUES ('10.219.238.86', '1130430104531000', 'Sche
dule', 'Processing Situation: kbk_schdclt_gskc_std')
2013-04-30 10:45:31.000000 Schedule: Situation Query Executing: SELECT SERVERNAME, SERVERVER, SERVERREL, SCHEDULENAME, NODENAME, SCHEDULESTART, ACTUALSTART, STATUS, RESULT, REASON, DOMAINNAME, NODE
TYPE, EXTRACTTIME from SCHEDULE WHERE EXTRACTTIME>'1130430095122000'
2013-04-30 10:45:31.000000 TSMDerbyDB: Error updating the DERBY database in updateDB


2013-04-30 10:45:31.000000 TSMDerbyDB: Message: An SQL data change is not permitted for a read-only connection, user or database.
2013-04-30 10:45:31.000000 TSMDerbyDB: Error Inserting message into the Database: INSERT INTO AGENTLOG (SERVERNAME, DATETIME, ATTRIBUTEGROUP, MSG) VALUES ('10.219.238.86', '1130430104531000', 'Sche
dule', 'De-Synchronizing DERBY')
2013-04-30 10:45:31.000000 Schedule: De-Synchronizing DERBY
2013-04-30 10:45:31.000000 TSMSchedule: No rows returned from DERBY query
2013-04-30 10:45:31.000000 TSMCustomClient: cp_WaitForRequest
2013-04-30 10:46:31.000000 TSMCustomClient: req.request is: 1
2013-04-30 10:46:31.000000 TSMCustomClient: this is a collectdata request
2013-04-30 10:46:31.000000 TSMCustomClient: Situation (non-historical-request) request found
2013-04-30 10:46:31.000000 TSMCustomClient: Situation name found is:kbk_schdsrv_gskc_std
2013-04-30 10:46:31.000000 TSMCustomClient: Situation name found is:kbk_schdsrv_gskc_std
2013-04-30 10:46:31.000000 TSMSchedule: Entering collectData as a Situation/Non-Historical Request
TSMcpci: Agent trace log will be stored in: /IBM/itm/logs
TSMcpci: Error converting KSK_KEYSTORE variable - going to default of blank


My tsm instance is admin of TSM application
I try to use user DB2 administrator (tsminst1) and user admin of TSM application >> both instance can view data on TEPS but situation monitoring cannot alert

please help me to find the way out. Many Thanks you all
 
PREDATAR Control23

Hi

I found the exact same thing... there is a bug that won't trigger a situation from a missed or failed schedule, even though the workspace shows the event.

I raised a call with IBM last week and eventually got a fix in the form of a new .jar file.. its now working fine.

Cheers
 
PREDATAR Control23

Hi

I found the exact same thing... there is a bug that won't trigger a situation from a missed or failed schedule, even though the workspace shows the event.

I raised a call with IBM last week and eventually got a fix in the form of a new .jar file.. its now working fine.

Cheers



Could you please share me a new.jar fixed please. I raised this issue to PMR support. They inform me to waiting new release. until now I not get any progress from them.
 
PREDATAR Control23

Could you please share me a new.jar fixed please. I raised this issue to PMR support. They inform me to waiting new release. until now I not get any progress from them.


this issue was resolve by load new TSMCustom jar file
who have the same issue please contact me via [email protected] to get new jar file and do following step

1. cd <install_dir>/itm
Example: cd /opt/tivoli/tsm/reporting/itm

2. Look for all of the TSMCustomClient.jar files.
Run this command: find <install_dir>/itm -name TSMCustomClient.j

Example find /opt/tivoli/tsm/reporting/itm -name
TSMCustomClient.jar

3. Find out which directory this file exist (different for each
platform and OS level)

4. stop all of the TSM instances <install_dir>/itm/bin/itmcmd agent
<instance_name> stop sk
Do this for all of the TSM instances

NOTE: you can also stop the instances from the GUI applicat
<install_dir>/itm/bin/CandleManage

5. Verify all of the TSM instances are stop by running
<install_dir>/itm/bin/cinfo -r

6. Make a backup copy of the TSMCustomClient.jar file
Once you have determined which directory the
TSMCustomClient.jar file exist in, then cd
into that directory
mv TSMCustomClient.jar TSMCustomClient.jar.MMDDYY

Where MM = Month DD=Day YY=Year

7. Now copy the new TSMCustomClient.jar file into this directory or
directories


8. restart all of the instances

<install_dir>/itm/bin/itmcmd agent -o <instance_name> start sk
Do this for all of the TSM instances

NOTE: you can also do this from the GUI application
CandleManage

9. Verify all of the TSM instances are started by running
<install_dir>/itm/bin/cinfo -r



Let me know how you get on



credit for Mr Chris Rees >> first hand person who help me solved this issue.
 
Top