Our blog guest is Norm Russell, Director at Automation Logic. A strategic partner to Hewlett Packard Enterprise, Automation Logicis a leading European professional services firm providing consultancy and support to large enterprises in the field of data center automation.
And here, we discuss all things data center automation - its challenges, the importance of metrics and culture shift, the best way to set up an automation team, what the future might hold, and a lot more …
SS: What comes to mind when you hear the term “data center automation”?
NR: Efficiency. Faster time to market. Building blocks for agility. It’s all about making things faster, and more efficient.
SS: In the context of infrastructure management, what are the challenges facing businesses today?
NR: Almost all of them are interested in delivering services faster, and much more efficiently. It’s sometimes about cost cutting, but not always. For example, traditional banks have to come to terms with competitive pressure from younger, agile, fin-tech digital banks who don’t have the burden of legacy systems. The barriers to entry are not what they used to be, and IT isn’t one of them anymore. Therefore, traditional companies need to rapidly rethink how they run their IT operations.
SS: At the onset of an automation project, what’s top-of-mind with Automation Logic clients?
NR: Our clients know they need to become more efficient and be more agile. But they realize that they are perhaps lacking the expertise to achieve these goals. They want to know how other companies have automated, and how automation is working out for them. They are concerned about painting themselves in a corner, having come from a more traditional, waterfall, and multi-year project approach. Hence, they simply want to minimize risks.
SS: What makes for a successful automation project?
NR: First, clients must identify the right metrics to appropriately measure business outcomes. Most large organizations have silo infrastructure domain teams, each with their own KPIs. However, standalone KPIs are almost irrelevant in the big picture: it doesn’t matter if one step in the middle is efficient, if the overall process is not rendered efficient. Metrics need to be holistic, and truly measure end-to-end improvement.
Then, of utmost importance, clients must embrace a culture shift. How clients run their IT operations has to change, and in some cases quite radically. In fact, the required culture shift is one of the biggest challenges to overcome, not always so easy, and often more difficult than the technical problems, because we are dealing with people and emotions. The culture shift requires a different way of thinking. It’s about changing the mindset and a realization of how value is measured. It’s about quickly determining if a product or a service, by way of automation, is adding value. And then, it’s about making the decision to proceed, pivot or stop.
SS: What’s one example of a really good success metric?
NR: Reduction in the amount of rework. It’s tied directly to high-quality, which ties to overall efficiency and cost avoidance.
SS: Specific to the data center, what functions have you helped clients automate?
NR: Some of the key functions are definitely around automation (of servers and databases) and orchestration of processes such as request fulfillment and service catalog management.
SS: What made you choose HPE Data Center Automation (DCA)?
NR: When it comes to IT operations orchestration, there really is no other product that is as good, as robust, and as flexible. And we have looked at numerous competitors, including open source. HPE operations orchestration does an absolutely amazing job tying disparate interfaces and processes across the data center. Then, it’s also an advantage that the HPE solution works across heterogeneous environments, particularly useful for large organizations who are laden with multi-vendor legacy apps.
SS: As a result of automation, what benefits have you helped clients achieve?
NR: Efficiency, measured in terms of faster time to market. With efficiency, two effects naturally ensue. First, cost avoidance, the case of “do more with less”. Second, improved quality achieved by way of eliminating manual hand-offs, reducing human errors and therefore rework. When software talks to software, quality improves immensely.
SS: What’s the best way to set up an automation team? Is it centralized, decentralized, neither, something in between?
NR: It really depends on where the organization finds itself in its automation maturity. Most organizations start with point automation, and they eventually form a centralized automation team. Whilst this structure can be more than perfect for small to medium sized businesses, large organizations start running into a problem of scale. In the face of finite resources, a backlog of automation requests start to amass. Hence, large organizations need to start thinking, and some already are, about how to enable and empower other teams to perform their own automation, by providing the right tools, best practices, and libraries of experience. Apple’s App Store uses this approach brilliantly. Apple has given the right set of guidelines and best practices to creative third-party software developers, and enabled them to be successful.
SS: If you could give one advice for a company starting their automation journey, what would you say?
NR: Don’t delay! In hindsight, clients often say, “We should have started automation a long time ago!” The cost of delaying is more expensive then jumping in, but jump in with the right mindset. Have the right mindset to create the right hypothesis and then test the hypothesis. Automation might not be 100% right the first time, but there’s always value in learning. As mentioned earlier, quickly make the decision to proceed, pivot or stop.
SS: Where do you see data center automation going next? What does the future hold?
NR: Bridging the gap between legacy apps and the newer, agile apps is one challenge organizations need to overcome in the next few years. It’s not uncommon for a typical bank to carry up to 10,000 legacy apps – trading apps, personal banking apps, retail banking apps, financial modelling apps… We did a quick study on how long it might take to replace these apps. It would take decades! Citing an example from a UK-only bank, less than 1% of the bank’s apps are cloud enabled and agile developed, despite a vigorous application replacement program it has been undertaking in the last few years. What’s a potential solution? It will require a new way of problem solving. Automation will play a big part in it, and will provide the right building blocks to enable the transformation.
SS: Thanks a lot for your insights Norm! One last question. First word that comes to your mind when I mention “data center automation and cloud”?
NR: Symbiotic! Automation is needed to enable the cloud and to use the cloud.