Tag Archives: 4-hour rolling average
We’re all being asked to warm up to IT climate change. With narrowing profit margins, companies put pressure on every department to reduce costs. IT, too long seen only as a ‘cost center,’ is particularly vulnerable. While hardware costs may be seen as more manageable, software and people costs look like better targets. In this ‘storm’ of cost-cutting, you want an umbrella or a shelter to protect you from the risk of a lightning strike that will eliminate your job.
In our last post, we reviewed the Automated Capacity Management (ACM) feature of ThruPut Manager and its ability to control the rolling 4-hour average (R4HA) by constraining or deferring specified workloads as the R4HA approaches the soft cap limit. But you may prefer more granular control in order to more fully leverage the opportunity for MLC savings, or you may not prefer to put caps in place at all, but still wish to reduce demand and enjoy the resulting cost savings.
Since the days when processor time was costly (and you input a job on punch cards), the CPU Busy metric has had intense focus. There are so many ways to look at the metric, all having vastly different meanings. Virtualization made it even more complicated. But for many in our field, this is still a very critical number. But is it the most important number?
When moving batch workloads around to lower your R4HA, duplicate product peaks are a common challenge—and can cause their fair share of headaches. To remedy this issue, IBM recently announced a new pricing model, Country Multiplex Pricing (CMP). The new model is designed to give you greater flexibility to move and run workloads across your data centers in a single country with less financial impact than you’d experience by staying with your present VWLC model.
Were you burned over the years by recommended capacity controls, such as hard capping or memory fencing? If you’re a long-time mainframe capacity planner, you quite possibly experienced the cost of implementing such ‘recommendations’—getting paged as key workloads were throttled and performance suffered. As IBM gets practical feedback from the field, it continues to offer better and better iterations of the tuning concept. Once an idea has been well field-tested and enhanced, it’s a good idea to take another look at it. Such is the case with soft-capping.
When it was first introduced, IBM’s sub-capacity pricing was a boon for capacity planners from a financial standpoint—allowing them to be more proactive in their planning. In the pre-sub-capacity era, all upgrades had to be carefully managed because of the huge potential impact on software pricing. Now, you can right-size your hardware and worry less about software costs—until you hit a soft cap, that is.
When it comes to lowering your company’s Rolling 4-Hour Average (R4HA), is capping always your best option? As we’ve mentioned elsewhere, while sub-capacity pricing can immediately reduce software costs, most installations need a way to guarantee a limit to the monthly charges to experience substantial savings. IBM offers a number of ways to do this, but virtually all of them involve the notion of “capping”. Capping is incredibly beneficial in certain scenarios; but, it isn’t necessarily the best choice for every installation.