Our methodology for managing Unisys Libra systems takes a top down approach and focuses specifically on online and batch processing capacity. If more detailed information is required, the RGS product can be used to browse information such as workload, program, unit (Pack, ODT, Printers etc), transaction and system information.
As part of any regular system review we would recommend Unisys Libra users use the System Summary Report as an initial view of how the system has been performing – this focuses on key performance metrics and allows decisions to be made rapidly concerning the state of the system(s).
In the following example you will see that the report contains data for just a single system over a period of time. Perhaps most importantly, this report allows us to look down the column titled “Online Utilisation” and make judgement regarding online utilisation. If all is okay then there’s little to worry about, but if there is something of concern, you’ll see it and later you'll see that action can be taken to discover why the problem occurred!
Click on the picture for the full image.
Of course that’s not the only use of the report though. Many of our users keep years worth of data which can be used to plot trends in online utilisation, transaction response time SLA's, transaction volumes, disk occupancy etc. Once armed with a sufficient amount of data other users have used this information to build forecasts of future levels of online utlisation, transaction/business volumes etc.
As mentioned earlier, we have to consider Unisys Libra systems from both the online and batch processing perspectives. Both are very different workloads and so we have to analyse them differently. In the example below, you will see a report that shows the flow of “critical” batch processing with total CPU utilisation mapped over the top.
Click on the picture for the full image.
This is a very powerful visual report that you can use to quickly identify any gaps in critical batch processing and make immediate savings if your month end batch processing over runs! Not everything in life is so simple though! Using information in the report you can start focusing your attentions on the “big hitters” and really begin to make an impact on reducing those batch elapsed times!
What about those daily batch processing runs? How long are they running for? how close to the start of online are they finishing? |