When I was asked to create a blog out here on SDN (there were several requests to do so), the expectation was that I would use the blog as an opportunity to bring interested readers a little bit closer to what is happening at the SAP Co-Innovation Lab. So the timing is good that I’ve recently made good on my commitment to get this blog underway right before our Virtualization Event. It seems like every few hours something comes out of the numerous team meetings and discussions that Roland and I have in planning for the week of April 20 that would be good topics for discussion on this blog, and then I get busy in COIL and fail to make the time to write. Then it’s the weekend where we aren’t supposed to let work invade our lives so easily but if you like what you do, then it’s nearly impossible so here I am on Sunday thinking about Virtualization Week in front of my blogging window. (I haven’t quite reached the point of all the planning keeping me awake at night or dreaming about it but this will probably change when we are only two days away from the event and not two weeks. Although, I think Roland might be be ahead of me in this respect!)
As for the topicality of this post, I’m thinking more about the content being covered as opposed to any of the logisitics supporting the event. One specific topic that I think will get good coverage from a few different perspectives is Disaster Recovery. Here in COIL we have pursued a very interesting and meaningful project with VMWare, NetApp, Intel, F5 and other partners across mulitple phases in both our Palo Alto and Tokyo COIL locations. The success of the project has really identified how innovation can be extracted from existing technologies and processes to effectively lower costs while also lowering risk is impressive and yet industrywide, we don’t exactly see executives lining up to sign off on budget for DR projects.
Rolling out mission and business critical systems without a DR safety net is risky but the fact is many of these systems are built upon dedicated server hardware making DR expensive.
We have several sessions on our agenda for the week that I am looking forward to. Session Ten on Day 2 from Slobodan Malbasic, and Session Fourteen on Day 2 from Steve Herrod will be two sessions that I am sure will illuminate some of major reasons for leveraging virtualization to address disaster recovery and compare what can and should be done today versus what alternatives will begin to present themselves through cloud computing. Relying upon the cloud of course will bring its own concerns from applying the right business model to issues of security. Certainly from a DR perspective, cloud computing may mitgate the additional hardware costs associated with providing the redundant systems used in disaster recovery architectures. What I am hoping these two sessions offer and perhaps from George Parker in Session Nine on Day 2, is some insight or quantitative arguments for how one builds a business case for disaster recovery.
Just over the last year based upon conversations I’ve had with peers in the industry and also from what I’ve read, many executives claim they understand the value of deploying disaster recovery in their data centers and yet they have yet to place it among their top IT priorities. This has perhaps become a bit like what it means to sufficiently invest in reducing network computing vulnerability- we are more inclined to take action after we are hacked. It may be that DR gets put off because if a server goes down, it results in unwanted downtime but that as long as the data can be restored from tape eventually, that this is sufficient. Maybe we have yet to fully quantify the performance hit we take when an end user is suddenly disrupted using a critical ERP application at quarter end or not being able to recognize quickly enough that a point of sale trasaction failure forced a customer to give up a shopping cart full of purchases to seek the products elsewhere. I’d like to see some of the sessions we participate in during the week to help all of us to better understand the importance and relevance of disaster recovery and demonstrate an economic rationale for pursuing virtualized disaster recovery solutions. If we also learn that there are still more problems to solve before the case can be made, then maybe we will see connections made at COIL that will result in new co-innovation projects being proposed and pursued to further advance this critical business process and the technologies that make it possible to support.