A No Holds Barred Conversation on Mainframe Batch Automation
Overview: Batch Automation Expert Kelly Vogt talks candidly about how companies are managing mainframe batch operations in the Digital Age. Read his enlightening Q&A and then check out his new white paper, “A New Batch Service Level.”
It’s a rare distinction to have a talented Operations leader join your company to help your customers address the very challenges he successfully overcame in his career. Such is the story with Compuware Solutions Consultant and ThruPut Manager expert Kelly Vogt .
Kelly has over 40 years of z/OS systems programming, performance tuning, capacity planning, operations and management experience. Prior to coming to work for Compuware, he managed an evolving batch strategy for a major insurer, improving batch services for his customers, while reducing operating costs and risks by fully automating batch processing.
Kelly recently penned a new white paper titled, “A New Batch Service Level.” We caught up with him to get his thoughts on how batch is evolving in the Digital Age and what he’s seeing in the field.
Q. Kelly, you write from experience. What kinds of things are you seeing at customer sites that inspired you to write this paper?
KV: Today, everyone is seemingly under a deadline. Everything is being consumed as a service and that includes batch processing. Gone are the days of chucking a job into the system and being happy to get it… whenever. People want transparency and consistency, and they are less and less acceptable of the mainframe as a black box.
DevOps has developers and testers working as rapidly as possible, running more batch than ever. Likewise, business departments that demand jobs when business milestones complete, are increasing pressure on batch, not to mention their regularly scheduled work. The operator has always been at everyone’s beck and call when job initiation is concerned. Now, it’s harder than ever.
Fortunately, ThruPut Manager’s ability to provide consistent batch service, with transparency via time-to-initiate and other historic service levels, enables the business to function within the tight constraints required today for profitability, whilst meeting customer requirements.
Q. Anything stand out in your mind that may seem internal to IT, but is actually impacting the business or customers on a broader scale?
KV: What I see causing the most harm is the constant threat of substituting imagined replacements, based upon the platform de jour, for efficient, scalable and reliable mainframe systems, for biased, uninformed reasons.
Q. Why has batch initiation performance become so complicated?
KV: The proliferation of job classes, coupled with the inherent time and cost pressures of batch operation leads to over-initiation, causing batch performance to suffer. Like a stopped clock, fixed initiator structures are the right size and shape twice a day, as the saying goes. And manual efforts to manage performance and cost often lead to vacillation exacerbating the problems. Human beings can only do a fair job, at best, in manually managing initiation levels for performance. Automation, making decisions based upon system performance metrics, is the only effective means to achieve consistently high performing batch.
Q. What level of batch maturity have organizations reached in your opinion?
KV: I wrote Compuware’s Mainframe Batch Maturity Model that consists of 5 levels of batch maturity: Reactive, Proactive, Automated, Optimized and Modernized. Many shops have taken at least some action to address shortcomings from the bottom up with job-related exits and system automation, eliminating much reactivity and making progress into proactivity. But few have made much headway into Automation, Optimization and Modernization where the real benefits are found.
Q. Do you think mainframe shops are prepared for the imminent retirement of their experienced batch operators, or are they taking a “wait and see” attitude?
KV: It depends on geography. In a few countries, mainframe shops have thoughtfully hired and developed talent internally. They realized the risk and are actively addressing it.
In the U.S., there’s a decades-long air bubble in the systems programming and operations talent pipelines. Few companies have an active plan. Many are thinking outsourcing, but many outsourcers have the same talent problems. Just because you outsource and make your operation the outsourcer’s problem doesn’t mean there won’t be blowback.
The thing to remember is automation reduces these risks dramatically. Using ThruPut Manager you can safeguard your domain knowledge by codifying it into our rules-based, policy-driven control system, instructing ThruPut Manager how to handle each job. And it can be evolved easily over time as needs present.