We systems programmers love metrics, and we have a lot of them. But in two areas, we struggle: SLAs and KPIs. Service Level Agreements (SLAs) always exist, even if you never agreed to any. The business and end users have expectations and, in their minds, these are SLAs. We may have set them for online but rarely have more than a “suggested” due-out time.
Key Performance Indicators
Key Performance Indicators (KPIs) are the kinds of metrics we can share with the business—ones that reflect a common understanding of a useful measurement, like the number of credit card transactions processed or the number of new accounts opened by day. The vast majority of the metrics we rely on are designed for IT users, not the business. They have meaning to those of us who understand all the multiple ways CPU busy can be used.
WLM doesn’t help much, and the metrics offered for batch don’t have the flexibility to allow a user to define how long they want a job to take. Yet, that’s precisely the kind of information he needs. Instead, you get a performance index, which is not intuitive to most people and doesn’t permit comparison between the expected due-out time and the actual.
As system programmers, we don’t generally communicate well with the business, which means business sees us more as a cost center than a value provider. I think we can do better. Perhaps we need to look for another metric, one that is useful in reaching across the aisle to the business. This is especially true in the area of batch performance.
A Brand New Metric
Within ThruPut Manager, we’ve created the Batch Service Metric (BSM) which allows you to define “target,” “acceptable,” and “critical” thresholds for each service group in your ThruPut Manager policy in real or actual wall clock times. ThruPut Manager does the best it can to deliver on the service required using queue management and job selection.
ThruPut Manager uses the BSM to measure and report on service. You can get immediate feedback on job performance including “in-flight” jobs and relative position in the queue. ThruPut Manager can even tell you when a job is going to start.
Leveraging the Batch Service Metric
But what really matters is how the BSM can facilitate communication. The business can define what they need in the same terms that Operations will use to measure the performance.
ThruPut Manager can help you set and achieve SLAs that are achievable and desirable. Give it a try and see how better communication between IT and the business can lead to career opportunities and advancement.