Skip to main content

A proactive approach to improving workload management

(Image credit: Image Credit: Rawpixel / Pixabay)

The push to be more productive and more efficient is a near universal goal of organisations across industries, of course, and there is no shortage of ideas on how this can be accomplished.  One way this can be tackled has to do with the way workload management is done. By strategically distributing work across an organisation’s workforce, workload management enables both organisations and their respective workforces to be more productive and efficient. When workload management is done well, employees maximise their skills and they’re more satisfied. However, if done poorly, bottlenecks take hold and employees – and their organisations – can easily become overwhelmed. Workload management has the potential to be a true business enabler but to realise this potential requires a new approach.

The fundamental need for workload management

One of the fundamental, critical processes in an Information Technology setup is workload management. That’s because batch systems prepare the business to run. For example, banks and insurance firms need to perform certain calculations by a certain time. Similarly, retailers need accurate daily processing inventory. They also need billing books and general ledger. Any delay or failure in workloads can seriously impact a brand’s image, in addition to potential financial losses.

One of the best methods for reducing a company’s costs is by using batch systems. This method can also increase employee efficiency. However, in practice, it has not always been executed well. That is at least partly because it has become so complex, with a hundred thousand or more jobs spread across business functions, multiple job schedulers and complex inter-dependencies. There is a good reason that the market for integrated workplace management solutions is growing quickly – it is clearly needed.

Trouble applying technology to the problem

Organisations are always in search of the latest solutions to solve significant problems, but they often operate without knowing how to apply their technology to solve their own problems. Consequently, the distance between business and IT increases, reaching a state where the gap becomes a technology debt. And that is a debt that only increases over time, even with the most modern technologies in operation.

Organisations must find a way to forecast how a workload system will perform and from there,  apply fixes before the problems occur. This is why workload management requires a cognitive approach – one that involves a technology-agnostic, comprehensive blueprint of the job streams. Such an approach would profile the normal behaviour analysis, coupled with a context-aware, self-triaging and self-healing mechanism.

Current workload management systems are problematic because they do not consider the batch data along with the various key performance indicators (KPIs) in the infrastructure based on a context-aware system. In an environment with changing jobs and dependencies, changing infrastructure and a changing business workload, a lack of context leads to a lack of end-to-end understanding. This leads to unexpected outages, inherently reactive operations and a process that is extremely difficult to predict.

Moving from a reactive to a proactive stance

Perhaps unsurprisingly, managing millions of batch jobs is a very complex undertaking. Diversified holiday calendars due to geographic spread, cross- and hierarchical dependencies, and a lack of automated performance metrics on the job scheduler are just some of the things that contribute to this situation. It gets worse when enterprises have multiple batch job scheduler solutions. In addition, the need for business to stay agile, relevant and creative in the competitive market introduces more than a thousand changes to the profiles of the batch jobs each week.

With the demanding nature of business driving changes in workload behaviour, it is quite clear how IT operations fall behind. Added to this is the fact that performance benchmark reports are becoming irrelevant in light of increasing technology debt. A singular focus on incident management is misguided, as well.

To ensure stable business operations and high-quality customer experience, timely and predictive processing of batch jobs is vital. Organisations need a new solution that helps move this from a reactive to a predictive approach, incorporating machine learning, artificial intelligence and automation to deliver agile and autonomous batch operations. This would enable the proactive fixing of issues before they occur and scenario planning for optimised batch runs.

Staying ahead of change

These days, business needs can turn on a dime. Workload management has tried to keep up, with vendors developing increasingly complex batch systems. Still, they cannot match the pace of change. Part of the issue is the technology debt that most organisations hold, which prevents them from fully meeting their needs. Artificial intelligence and automation offer a reset, particularly as they demonstrate their abilities across many business applications. Organisations looking to future-proof their workloads will use the capabilities and benefits of these technologies for their batch jobs. When they do this, they will serve all stakeholders better by implementing workload management that is predictable. And that means avoiding downtime and guesswork.

Akhilesh Tripathi, global head, Digitate

Akhilesh Tripathi is global head of Digitate ( a software venture of Tata Consultancy Services. He was previously the head of TCS Canada, where he drove the Canadian entity to be among the top 10 IT services company in its market.