When servers appear to be missing from the the availability reports it could be because of many reasons, but whenever I run into this type of issues these are the kinds of troubleshooting steps I take.
First of all, to troubleshoot this issue we should be clear on the workflow going on behind the scenes that allows these reports to work:
- The agent computer generates 7036 events in the system log for the start and stop of services.
- If the computer is configured as part of the computer group associated to the availability reporting management pack, these events are collected by the MOM agent to the Onepoint database.
- When the System Center reporting DTS task runs (once per day) the data is transferred from the Onepoint database to the System Center reporting database (to the SC_* tables).
- The Reliability Availability Scheduled task runs, transferring this data to the MRAS_* tables within the System Center reporting database. By default this runs every 2 hours.
- If a report is scheduled, the report is generated at the time closest to the scheduled time.
If any of these fail then the report will not be generated.
A few of the issues are listed below:
The system is not generating 7036 events:
The 7036 events are generated by the System Control Manager. In case they are not getting generated for a stop or start of the service try the following:
- Make sure that EnableAutoRestartEvents = 1 under HKEY_LOCAL_MACHINE\System\CurrentControlSet\Control\
- Recompile the SCM MOF file by running the following command:
- If compiling fails, make sure that the scm.mof file is copied from a working system and placed in the location C:\WINDOWS\system32\wbem\and then try compiling it again.
The system is generating 7036 events but some servers are missing from the reports:
- Make sure that the system is added as a member of any computer group that is associated with the Microsoft Availability Management pack.
- If the system is not appearing on any specific report like DC reports or Exchange reports make sure the computer groups are associated to the corresponding elements inside the Microsoft Availability Management pack.
- Check if the rules are enabled by running RSOR.exe <DBSERVER> <AGENT>. The tool RSOR stands for Resultant Set Of Rules that is a part of MOM tool kit. The results are displayed in C:\resultantsetofrules. Make sure that the following are enabled:
· Event – Run Availability Reporting Service Discovery (Enabled)
· Collection – Availability Reporting Service Synthetic Uptime Event Collection (Enabled)
· Event – Availability Reporting Service Synthetic Uptime Event Generation (Enabled)
- If you are looking for a new system which has recently been added, you need to make sure that these rules have been run. By default they run every 23 minutes. Also make sure that the system has been rebooted at least once after the system had been added as a part of the computer group associated to the management pack.
- In order to check if the data for a particular system is being received, please runs this query against the System Center reporting database:
SELECT distinct ms.nvcServerName,*
FROM mras_Sojourn s JOIN MRAS_ApplicationVersion av
ON s.ServerVersion = av.uidApplicationVersion
JOIN mras_server ms
WHERE modelname like ‘%<ModelName>%’
ORDER BY timein desc
Note: <Modelname> is Exchange ,AD,SQL,OS etc and <Servername> is the name of the server not appearing in the list.
- Check the output of the query above to see when the last data has arrived for that particular system. In case the system does not appear, please check from the first step above to help identify the cause. In case the data is being received on the server but it’s not showing up in reports, make sure that the scheduled report has the corresponding computer group and domain correctly configured.
- Make sure that both scheduled tasks run successfully. If the System Center reporting DTS task fails, check the following articles for more information on troubleshooting the issue:
To troubleshoot ReliabilityAnalysisReporting you can use:
- Make sure that the scheduled report has run successfully. To check this run the following query:
Select * from dbo.MRAS_ReportFriendlyname where reportName like ;%<ReportName>%’
The report name is the name of the report given when we scheduled the report. From the output of this query you will get the uidreport value which corresponds to the report we are looking for. With that value of the uidreport run this query:
select * from dbo.MRAS_ReportRunLog where uidreport =’<the result of the previous query>’
Then check the value of the istatus column corresponding to the report:
0 : Success
1 : Failed
2 : No Data
- Check if there are multiple entries for the system in the database. To do this run the following query (The first one against one point and the other two against system center reporting):
· select * from dbo.computer where name=’<SERVERNAME>‘
· select * from DBO.SC_COMPUTERDIMENSION_TABLE WHERE computername_pk= ‘<SERVERNAME>‘
· select * FROM DBO.MRAS_SERVER WHERE nvcServerName=’<SERVERNAME>‘
- If you find that there are multiple entries make sure that the following are done:
- Delete the Computer from the MOM Administrator console after moving the agent to the Unmanaged Computers node.
- Delete the extra entries if any of these tables show rows related to the deleted computers (dbo.computer, DBO.SC_COMPUTERDIMENSION_TABLE, DBO.MRAS_SERVER)
- Reinstall the agent on the computer and reboot it the next day. Let the two scheduled tasks run afterwards.
Hope this helps,