Question & Answer
Question
When opening a new PMR with the MDM support team, it is best to have the following information available so that the support team can easily review logs and information for the issue. The first table, All PMR issues, is information that should be gathered for all PMRs. The other three contain useful information depending on the issue that is being encountered. Please review the tables below and have this information available before opening the PMR.
Answer
All PMR issues | |||
| Data | Details | How to? | Type |
| Business impact | MUST COLLECT | ||
| Complete Stack Information | MDM version Database type and version including patch levels, OS, appserver, DB server | For DB2, you can use db2level, for Oracle, use SQLPlus Login. | MUST COLLECT |
| Is the issue happening in production or on a development/test system? | If it is on production, are you able to reproduce the issue on a test/dev system also? | MUST COLLECT | |
| Recent changes to the system | Did the problem appear after recent changes? | If there have been recent changes, what were they? | MUST COLLECT |
| Are the appserver and db running on the same machine or different? | If they are on different machines, please provide the stack info for both boxes | OPTIONAL | |
| Is this issue reproducible? | If the issue is reproducible, please stop the MDM server, clear the logs, restart MDM, reproduce the error and note the timestamp. Then please send the complete WAS and MDM logs. | MUST COLLECT | |
Performance specific issues | |||
| Data | Details | How to? | Type |
| enable performance tracking | perfConfig.sql | log4j.appender.performance Log.File=filename.log For more detail information, please refer to Chapter 17. Tracking performance in MDMDevelopersGuide.pdf for MDM | OPTIONAL |
| IBM MDM performance collecting data document |  IBM_MDM_Performance.pdf | Please refer to the document. | OPTIONAL |
| monitor process list using # ps -aux | This will give support an idea as to what else on the system might be causing the performance problem | put the script into cron.hourly paying attention to the cpu%'age used by each process, count process loads and look to see how many non-MDM services might be using up the CPU time | OPTIONAL |
| Any changes to the hardware/OS | Have there been any changes to the system hardware or patches to the OS? | OPTIONAL | |
Out of memory issues | |||
| Data | Details | How to? | Type |
| monitor process list using # ps -aux | This will give support an idea as to what else on the system might be causing the performance problem | put the script into cron.hourly paying attention to the cpu%'age used by each process, count process loads and look to see how many non-MDM services might be using up the CPU time | OPTIONAL |
| OS running out of memory - Getting full coredump on AIX platform | How to produce a full AIX core dump | OPTIONAL | |
DB Specific issues | |||
| Data | Details | How to? | Type |
| Recent changes? | Are there any recent changes/patches to Database or server OS? | OPTIONAL | |
| Query performance | Has the DBA checked for segment fragmentation/reorg check recently? Without this, you may see slow performance of the app | DBA should have maintenance scheduling | OPTIONAL |
| Server resources | Are there any abnormal resource utilization in db server? | Use vmstat for quick overview, iostat to check filesystem load, ps -aux for detailed process and thread resources | OPTIONAL |
| collect db information | For Oracle, run the following query: select * from v$parameters For DB2: run db2support utility | OPTIONAL | |
[{"Product":{"code":"SSPVUA","label":"IBM InfoSphere Master Data Management Server"},"Business Unit":{"code":"BU059","label":"IBM Software w\/o TPS"},"Component":"General product information","Platform":[{"code":"PF002","label":"AIX"},{"code":"PF027","label":"Solaris"},{"code":"PF016","label":"Linux"}],"Version":"8.0.1;8.5","Edition":"","Line of Business":{"code":"LOB10","label":"Data and AI"}}]
Was this topic helpful?
Document Information
Modified date:
16 June 2018
UID
swg21370106