This week at HP Discover, we had an interesting panel discussion about the future of mission-critical computing. The panel consisted of industry analyst, customer, Intel and HP execs discussing topics ranging from how the market defines mission-critical today to trends that will impact the future of Odyssey, a project to redefine the future of mission-critical computing by cascading key HP-UX functionalities to Linux and Windows platforms. Here are some highlights I took away from the panel discussion:
Defining mission critical
Is the definition of mission-critical computing evolving? What does it really mean today? Is it at the infrastructure vs the application level in how you define it?
- The definition can be at both the Infrastructure and the application level. Generally the nature of the application and the real business impact of any downtime determines what is and isn't mission-critical. It is going to vary industry by industry and company by company to some extent.
- Consider truck delivery of supplies that show up during the lunch rush – this would be simply unacceptable. So if it disrupts the customer and you could risk losing that customer, that would also be considered mission critical
- It can also have to do with the mission of the company. If the system goes down, does it require CEO/Board approval? A financial disclosure?
- As it applies to services, high touch, high sense of urgency are what matters in mission-critical situations. The right experts must be available immediately to address the situation and resolve the issue correctly, immediately. In services, it is all about business outcomes and the experience and ensuring long term relationships are protected, honored and preserved
What do people expect from mission-critical infrastructure? Where is it heading?
- HP has heard from customers that they ultimately want the Superdome mission-critical experience in x86
- Panel notes there are still critical differences at the platform level between Xeon and Itanium. There is a path of development ahead to deliver more on Xeon. The good news is it is one unified team designing the convergence.
- A note of caution to CIOs/IT architects - even if your extremely intelligent CEO tells you to do it, it doesn’t necessarily mean it is the best path for mission-critical. There is still a need to have a system that is truly ready, really there to handle mission-critical. Evaluate carefully the full system, services and support you will receive throughout the entire lifecycle of your investment before you leap.
- Also be sure to consider the transition of mission-critical technology and the people with the skills to support it. Will there be an experience gap? Will your vendor provide common tools to help manage the system now and in future? Is the system designed to carry forward for the next series of decades with the skills and preferences of your people in mind?
Trends Impacting Mission-Critical Computing
- Customers are putting more workloads on x86
- Customers are desiring a single point of contact for mission-critical services as hybrid experiences/platforms develop in organizations
- Explosion of mobility. Applicationss have to be up 24/7. Applications are increasingly considered mission-critical and cannot go down. End users will expect to access mission-critical applications from their choice of mobile devices.
- Cloud is an increasingly popular consideration. It is recommended that as you work with saas vendors, you also evaluate their infrastructure to deliver. The first time you have a 2 hour outage you really ask questions.
HP, Intel mission-critical landscape
Three years or so from now, what will the mission-critical landscape look like? What is HP working on? Where is Intel focussed?
There is an Interesting tension arising – mission-critical systems are typically measured in decades. Now there is external pressure to move at more rapid rate. However the world does not change overnight. It really isn't about flash cuts in mission-critical environments. Co-existence and convergence is the more logical and appropriate path.
From Intel's perspective, the products are already in the works such as the Poulson chip. They see continuing convergence in x86 and itanium. Intel will bring the right Itanium qualities to Xeon. ll about coexistence and convergence. The world does not change overnight. We will live with and coexist. It will not just be a rip and replace. We will endeavor to converge from a trend perspective.
- At HP, we believe converged infrastructure is the best path. We will continue to have one infrastructure for OpenVMS, NonStop and Integrity. We will use this same infrastructure for x86 Linux and Windows. We will thoughtfully develop and deliver the mission-critical experience in x86 that customers expect and deserve. It will be a journey, and we are working with our services team hand in hand as we develop this path with Project Odyssey to ensure our customers get a true mission-critical experience.
- HP will respect your timeline. If you want to modernize one application at a time, one workload at a time, one site at a time we respect that. We want to help you achieve whatever works for YOU at your pace. Convergence is key. Common platform is key. Going at your own pace is key. If it takes a decade that is ok. We will be with you through the journey.
Thanks to all of the panel participants for a very interesting, lively discussion.