So, here we are post-Hitachi NEXT 2018 and the brain and body are just starting to recover. There were certainly a lot of interest and buzz around numerous topics during the week, including IoT, NVMe and All-Flash. However, the topic that
seemed to take center stage and draw lots of interest amongst the attendees was around AI and Automation.
Why the rise in interest for AI and Automation you ask? Well, a couple of thoughts come to mind: The first, AI continues to increase in data center use as companies are seeking to leverage advanced correlation and analysis that can be obtained from AI. The second, IT organizations are beginning to look at combining AI derived analytics with automation as the next logical phase in helping them to create a new IT operations model; referred to in the market as AI Operations.
Now, even though interest levels are rising, most customers I talk to are still in the infancy stage of bringing automation into their IT operations. Although it’s obvious that automation will bring numerous benefits, most IT organizations are struggling with how to best approach it. As IT automation continues to evolve, it raises many questions that IT organizations must address including:
- What could the benefits that automation brings to IT organizations?
- What are the biggest challenges that automation raises?
- Are these challenges technical, cultural, or both?
- How will the evolution of automation change the future of IT operations?
To address these questions, I recently sat down with Paul Lewis, Global Vice President and CTO, Industry and Enterprise Architecture, Hitachi Vantara. Paul joined me on a panel discussion at Hitachi Next 2018 that focused on the topic of AI and Automation (and by the way, I just learned that it drew the largest attendance of all the breakout sessions). In his role, Paul is in constant contact with executives, partners, and influencers discussing technology evolution, challenges and what is driving their success. Paul and I will be addressing similar questions that we discussed during our session at NEXT, so let’s begin the discussion:
Stan: Hi Paul, why do you think IT organizations are looking at AI and automation?
Paul: Well, automation is required more so today because of diversification in the data center. With increased IT challenges around acquisitions, increased adoption of cloud and SaaS proliferating across data centers; this has
led to further infrastructure silos and inefficiencies in the IT organization.
Couple this with years of insufficient system updates and standardization programs that led to several versions of the same software product on the floor; IT organization are facing resource challenges to manage incoming technologies, let alone clean up version problems.
Stan: So, help our readers understand what you’re hearing from customers as to what’s driving their interest level in Automation?
Paul: Automation exists to bridge the gap between the digital and physical world of data center management, from allocation of storage to monitoring HVAC units. Automation extends the analysis that customers can gain from leveraging AI; in collecting and correlating the complex data gathered across their physical, heterogeneous infrastructure platforms.
Automation is a natural extension to AI in that it “automates” the “reaction” portion that is derived from AI analysis of the physical platform. In short, AI helps to filter and report real-time alerts that automation can react to “automatically” and it provides the predictive analysis that is vital for longer-term planning.
Automation, and what it can bring to efficient IT operations, is what is creating the interest and excitement in the market especially as it continues to evolve.
Stan: Well, this sounds all well and good but are there obstacles that are hampering customers from moving forward with automation?
Paul: As it was with virtualization, the IT organizational structure tends to be the initial roadblock to automation adoption. When virtualization was introduced, it obviously cut across the compute, storage, and networking; which in the traditional IT organizational structure were operating as separate teams and with different technical disciplines.
Eventually, IT organization realized that they needed to create virtualization teams; which could leverage virtualization based on workloads. These teams would support the virtualized workloads and work through the internal politics and resistance both inside and outside of the IT organization.
Automation crosses these same data center groups (compute, storage, networking), as well as IT operations, ITSM and release management. Therefore, its paramount that companies look at creating Automation teams; develop automation skills and create specific roles such as automation engineers or automation architects.
Stan: OK, so then what do you think is the most compelling opportunity for AI and Automation today?
Paul: Driving business outcomes is still a top priority for IT organizations as they are chartered with helping their business adapt quicker to market shifts; and business outcomes has less to do with better management of infrastructure or faster time to recovery. It’s about better analytics, models and metrics that help the IT organization with better planning.
In the past, IT would have to "stuff in" base infrastructure changes into random projects, which resulted in cost overages that had a ripple effect across the organization. With AI based automation, you will be able to better predict what growth or changes you’ll need to make to your data center, by quarter; if you need to increase capacity, where additional virtual machines are required, and what workloads are actually taxing the data center. Automation greatly speeds up the “resolve” as it can automate the spinning up of additional storage and can deploy additional virtual machines that best meet specific workload requirements.
Stan: So, to close our discussion out please share with our readers what you believe is the future of automation is and where can it take us?
Paul: What does the future hold? Well I think that the concept of “lights out” data centers, where all operations are automated, and limited human interaction is only required to address predictable physical maintenance of the hardware; is certainly closer to becoming a reality than before, especially with the continued evolution of IT automation.
Think about it from this perspective, how customers interact with cloud companies such as AWS or Azure today will be very similar to how they interact with their own IT organization moving forward; submit request for additional workload or capacity, the service is deployed, the invoice arrives in the mail. Except for the invoice part, the process will be very similar moving forward.
What got left out! Some additional words of wisdom
As a former Industry Analyst, I feel that it’s my civic duty to interject some additional insights that I’ve gleaned from conversations with customers and industry experts on the topic of AI, Automation, and AI Operations. Although the focus of this blog was mainly on AI and Automation, the term AI Operations always tends to get woven into the conversation. So here are some things to remember as you proceed forward.
- AI Operations is still a relevantly new concept (yes, not a product) that was introduced (and defined) by Gartner Research only a few years ago.
- AI Ops could be confusing to understand; many different interpretations, and many moving AL/ML parts including Monitoring, Service Management, and Automation.
- The key really is to not get too caught up on the concept of AI Operations, and focus on what’s important from an IT
Operations perspective; which is how to leverage AI & ML and how and when do you weave in IT Automation.
Now with respective to IT Automation, note that it’s going to be a journey and not a sprint. Having an automation strategy with a phased approach is likely the best path forward. Although automation can be viewed as a “scary” or bad term to IT 0rganizations, automation will provide IT with the opportunity to increase their value to the business by focusing on more
strategic initiatives. As IT Automation continues to evolve, increased support for tight integration with existing management frameworks and applications, such as service management (i.e. ServiceNow) and open standards (i.e. OpenStack, REST APIs, etc.) will be paramount.
How can Hitachi Vantara help you through your journey to AI Operations:
- Hitachi Vantara is a leader in Data Center Modernization and offers a comprehensive portfolio of storage arrays.
- Hitachi Vantara provides a full-compliment of AI Operations software solutions through Hitachi Infrastructure Analytics Advisor and Hitachi Automation Director.
- Analytics is blending with the data center infrastructure, partnering with a vendor that has expertise could be vital to the success of your IT transformation.
- Providing open API’s that are easily leveraged by Hitachi Vantara partners; which extend Analytics and Automation
(ex: ServiceNow integration with Hitachi Automation Director).
Expect to see more blogs on this topic in the future as there are lots more that I can cover. Until next time!