I haven't looped back to this thought in quite a while, but felt motivated to do so by this recent press release announcing EMC's business continuity professional services for virtualized environments.
Once again, it's another example of virtualization changing everything, including how we think about protecting against the unthinkable.
The Problem(s) With Remote Business Continuity
Having the privilege of talking with many customers about this topic over my last decade+ at EMC, I usually boil it down to a few structural challenges.
First, there's the psychological barrier.
Some businesses take business continuity seriously, others don't. Rather than praise the ones that do (and rant at the ones that don't) I've simply come to accept that -- like buying life insurance -- it's one of those decisions we tend to make as we mature and have more to protect.
Put differently, I can rant about the need for life insurance to my 19 year old daughter, but I doubt that she'd see it as a priority until there was more at stake.
Second, there's the cost barrier. This can be expensive stuff to put in place.
Fortunately, the hard costs associated with a remote business continuity capability have fallen by orders of magnitude over the last few years. Networks are far cheaper, replication technology is far more efficient and flexible, and virtualized servers require far less kit to provide protection than their physical counterparts. And, yes, VMware is a big part of this cost reducer.
Third, there's the operational aspect.
Remote business continuity environments have to be monitored to make sure their always ready, they need to be tested occasionally to make sure they'll work as advertised, and "configuration drift" has to be mitigated to make sure changes aren't made that degrade (or break) all that protection you put in place.
Operational support means people, which means money. VMware-based environments using SRM (site recovery manager) have proven themselves inherently easier to monitor, and go a long way to controlling config drift through encapsulation and use of templates.
The big win here is testing -- using virtual machines, it's possible to test business continuity capabilities in the middle of a normal workday, if you choose. For those of us that treasure our nights and weekends, that's a huge plus. And, of course, SRM makes it all straightforward.
Finally, there's the planning piece. DR and BC pros will tell you it's all about assessment (what needs to be protected, and why), design (you'll be living with this for a long time), implementation (a fairly broad range of skills are required across the technology stack), building the operational plan, and -- ideally -- going beyond simply recovering applications to recovering your business.
The Rationale For EMC Services
Look, when people invest in remote business continuity, they're essentially investing in insurance. You're buying protection. You want to make sure you get what you need, it works well, and get it at a good price.
And, whatever you thought you knew about this topic in the physical world, many aspects are fundamentally different in the virtualized world. New and attractive approaches are possible -- and should be considered.
I suppose that the other important part here is experience. I have no way of proving it, but I firmly believe that EMC technology powers the vast majority of enterprise-class remote business continuity environments to this day.
It's a market we essentially created with SRDF way back when (1995?) and have continued to extend with complementary technologies since then.
Sometimes we get criticized for having so many products that essentially replicate and protect data. I usually counter that I probably wouldn't want a surgeon who had only one scalpel to work with.
Finally, there's new flexibility in the consumption model for this sort of stuff. The "managed services" component is proving very popular indeed with IT shops that need the protection, but don't want to invest in the operational staff and process required.
I think we'll see more of this in the future.
And Then There's The Private Cloud Angle
You knew I'd sneak this in again, didn't you?
As we really consider fully virtualized environments, there are some newer scenarios that are gaining traction that are worth putting on the table.
One of the recent ones is the "reverse DR" scenario -- hand all your production workloads to a compatible outsourcer or service provider, and use your own facilities as your own DR site. Keep enough on-premise resource to do development, testing, qualification, integration, etc. -- but put the operational loads in a purpose-built share data center facility.
And, of course, doing this in a virtual environment is far easier and cheaper than a physical one :-)
I don't want to get too far ahead of myself here, but it's logical to forsee that workload balancing between sites and traditional DR approaches begin to converge in the future. Sure, we see it today in uber-high-end mainframe environments, but it won't be too long before we're doing the same things with virtualized environments on industry-standard hardware.
Again, a useful theme to add into the planning scenario for consideration.
Once Again -- Virtualization Changes Everything
OK, I was out there pretty early, ranting and raving how virtualization technology (specifically VMware) was game-changing in ways that were hard to foresee.
With this announcement, I guess we've added another proof point to the discussion!
Indeed it changes everything..
if you want to hear more IEC about how we helpded them in this project(including best practices and how to avoid pitfalls), come and join us at VMworld 2009 at the following session:
BC2537 - VMware Site Recovery Manager with Enterprise Replication - A Customer Testimonial
it will be a blast
Itzik
Posted by: Itzik Reich | June 18, 2009 at 02:51 PM
Virtualization also breaks everything.
You need to backup differently.
You need to protect differently.
You need to manage differently.
You need to secure differently.
Hence you have a large ecosystem of startups building new solutions for the Virtualized Data Center.
Posted by: terry | June 19, 2009 at 10:07 AM
Hi Terry -- I was sort of with you until you pointed to startups as the answer.
I can't speak for others, but I can point to a few billion dollars of M+A, R+D and integration work that EMC has invested directly along these lines.
We don't have a monopoly on good ideas, so there's room for many to play -- but you can't argue that this sort of mega-investment isn't entirely a bad play :-)
Posted by: Chuck Hollis | June 19, 2009 at 01:42 PM
A really interesting green computer technology I found is Userful Multiplier. It's where multiple people can use the same computer at the same time each with their own monitor, mouse and keyboard. This saves a lot of electricity and e-waste. A company called Userful recently set a virtualization world record by delivering over 350,000 virtual desktops to schools in Brazil. They have a free 2-user version for home use too. Check it out: userful.com
Posted by: bj79 | June 22, 2009 at 12:45 PM
Good points!
You might also want to also check out the recent disaster management posts on BCP NEws - www.bcpnews.com.
Posted by: Gaston Boisson, CBCP | September 16, 2009 at 03:36 PM