In the first part of our look at integrating IBM Tivoli Workload Scheduler (TWS) and IBM Tivoli Monitoring (ITM), we will be focusing on the integration between TWS 8.4 and ITM 6.1. In later installments we'll be looking at replicating the same functionality within ITM 6.2, and also we'll be covering some advanced topics for improving the visibility of your TWS events inside the Tivoli Enterprise Portal (TEP) over and above what the out-of-the-box solution provides.
Integrating TWS 8.4 and ITM 6.1 - The Basics
If you've forwarded events from TWS to event management products like Tivoli Enterprise Console in the past, or have used the IBM Tivoli Monitoring v5 Resource Models to look after the health of your TWS agents previously, then the ITM v6.1 solution will be easy enough to pick up. It is expected that you are familiar with the workings of ITM 6's Tivoli Enterprise Portal, and have a basic understanding of the function of the Universal Agent to proceed.
There are 3 components involved in the TWS 8.4 to ITM 6.1 integration - a Universal Agent script provider, a Universal Agent file data provider, and some customsed TEPS workspaces.
Script Data Provider
In the past, you may have used IBM Tivoli Monitoring v5 resources models to monitor the TWS processes on the agents throughout your network ensuring critical processes like netman, batchman etc were available. There were also resource models that checked on things like disk space usage and swap space availability, for example. Any threshold breaches would normally be forwarded up to a Tivoli Enterprise Console server for further action. ITM 6 offers similar functionality to those from the ITM 5 resource models but with some alerts to monitor things like the amount of space used by the stdlist and schedlog directories, and others that check the size of the all-important message files (.msg), or ones to look after the new processes for TWS 8.4, such as monman.
File Data Provider
The file data provider of the Universal Agent (UA) works in a very similar way to the method of old where you would configure BmEvents.conf to have batchman write out the TWS events of your choice to a file, which would then be parsed and formatted by a TEC logfile adapter, and forwarded onto a TEC server for further action. The ITM 6 UA works in the same manner, and is configured by a product-provided script, with the events being displayed in the TEP, rather than a TEC console. Out-of-the-box, it will only display 'critical' jobs in the TEP, but this can be modified to have any event of your choice alerted upon within ITM 6.
Monitoring Your TWS Environment
As discussed above, the overall health of your TWS environment is looked after by a script supplied with TWS 8.4 which is located in $TWSHOME/ITM.
Directory of C:IBMTWS ws84ITM
14/07/2008 20:39 <DIR> .
14/07/2008 20:39 <DIR> ..
17/03/2008 12:14 970 baseNameSituation.xml
17/03/2008 12:14 1,976 ITMconfig.cmd
17/03/2008 12:14 11,378 ITMCreateSituations.cmd
17/03/2008 12:14 3,602 ITM_TWS_84metafile_win.mdl
14/07/2008 20:39 <DIR> scripts
4 File(s) 17,926 bytes
3 Dir(s) 2,946,228,224 bytes free
Install the script by running ITMconfig.cmd and pointing it at your ITM home directory.
C:IBMTWS ws84ITM>ITMconfig.cmd -ua_home_dir c:IBMITM
KUMPS001I Console input accepted.
KUMPS020I Import successfully completed for C:IBMTWS ws84ITM..ITMITM_TWS_84metafile_win.mdl
KUICRA007I: Restarting UM agent(s).
KUICRA015I: All UM agents on the local system were restarted.
The IBM Tivoli Monitoring 6.1.5 integration has been successfully enabled.
There are a set of supplied situations that come as part of the integration but care should be taken when installing them into a production environment. Every provided situation has a severity of CRITICAL, and as there are situations for both an UP and DOWN state, several of them will fire as soon as they're installed. They won't create events on the Situation Event Console if they haven't been associated with a navigator item, but it's something you should look out for if installing from the defaults. Another gotcha is that the metafile that's used to parse events into ITM uses the 'sampled' method for some of the monitors so you'll often have both the UP and DOWN situations firing at the same time. A workaround for this is to amend the metafile to use the 'polled' method instead of sampled.
C:IBMTWS wsadminITM>ITMCreateSituations.cmd -user sysadmin -password mypassword -host dellw2k3 -ITMHome c:IBMITM -TWSHome c:IBMTWS wsadmin
KUIC00007I: User sysadmin logged into server on https://dellw2k3:1147.
KUICCS010I: The situation TWSBaseName was created on the server on https://dellw2k3:1147.
KUICCS005I: The situation TWS_SU_schedlog_Critical was created on the server on https://dellw2k3:1147.
KUICCS005I: The situation TWS_SU_schedlog_Warning was created on the server on https://dellw2k3:1147.
Click to enlarge
Monitoring Your Jobs and Schedules
To monitor your jobs and schedules using ITM 6, IBM provides a replacement to using a TEC LFA to monitor the output of batchman on the MDM. The support files are located in the INTEGRATION directory within the TWS fixpacks section of IBM's FTP site. The last update at the time of writing was at the link below.
After unpacking the tar, head to the TEPClientConfigCD_ROOT directory and run setupwin32.exe which will create some workspaces and add some product support to your TEMS and TEPS.
Now install the file data provider metafile using the supplied script.
Run C:IBMTWS wsadminTEPTEPconfig.cmd -ua_home_dir c:ibmitm -eventfile_path c:ibm ws wsadminmevent.log -metafile_path c:IBM ws wsadmin epTWS84metafile_Sample.mdl
The TEPconfig script will copy BmEvents.conf to the TWS home directory and will amend it so Batchman will write it's events to the location you chose above (I use TWSHOMEmevent.log). Once batchman has been restarted, any of the events you specified in BmEvents.conf will be parsed by the UA and displayed in the TEPS.
Using the out-of-the-box configuration, only critical jobs and jobstreams are monitored, and no situations are provided.
Some basic workspaces can be put together quickly to give you an overview of your TWS processes across the network alongside workload throughput and jobs/schedule alerts. Any output can be warehoused for reporting and tracking purposes.
To summarise, the out-of-the-box confguration for integrating TWS 8.4 with ITM 6.1 can replace the functionality provided in the past with ITM 5 and Tivoli Enterprise Console. To realise a fully-functional integration, extra steps are required to customise the output and to build meaningful workspaces, situations and queries that will give you the most benefit from combining the two.
In the next article we will discuss integrating ITM 6.2 and TWS 8.4, and TWS 8.3 for those people yet to upgrade.