I regularly visit numerous public and private sector organizations in my role at Compugen and I continue to see that many still aren't addressing business continuity (BC) with any real understanding or commitment to a contemporary roadmap that aligns with today's new thinking and new technologies. In big-picture terms, that roadmap, which of course will also be shaped by circumstance and budget, looks something like this–and recommended in more or less this sequence:
- Identify critical apps and establish RTO/RPO objectives
- Re-centralize branch infrastructure to primary datacentre
- Virtualize /consolidate your server environment
- Build high-availability into your IT environment locally
- Establish a secondary datacentre site (internal or third party)
- Deploy automated failover technology (e.g., scripting)
- Optimize the pipe linking primary and secondary datacentres
Let me use this blog, and several subsequent postings, to offer some personal thoughts and observations regarding what's happening out there for organizations at critical steps along this roadmap.
- On RTO/RPO objectives … no profound insight here, but it's still hard to know 'how' to get somewhere unless you know where you're going; and if the BC goal is "five-nines" (i.e., 99.999% uptime), then you need to know this upfront since it will significantly shape your decision-making all along.
- However, your budget may never allow you to completely protect everything against possible disruption; but then again, you probably don't need to. You likely have low-priority applications and business processes that can easily withstand a little downtime in the event of an outage or disaster … do I smell "80/20 rule" here?
- On server virtualization & consolidation … here's one step on the roadmap where many organizations are still coming up short in terms of logical and effective BC practices. Many firms are still at a low 15% to 25% server virtualization; many are still afraid that their critical applications won't run on a hypervisor–folks, it's a myth and they'll all work; and still others ask "why virtualize a complex system such as Exchange Server when there may only be a one-to-one physical-to-virtual mapping and thus no apparent server consolidation benefit?" Those asking, clearly aren't seeing beyond consolidation–one of the most basic attributes of running on a hypervisor is that it will also give them quick and easy 'server replication', even to unlike hosts, and thus the ability to support high-availability locally or remotely.
In subsequent posts, I'll comment on the rest of the BC roadmap. In the meantime, feel free to e-mail me if you have thoughts on BC/DR you'd like to share, or to find out how Compugen expertise, solutions and services in this area can help you ensure five-nines of uptime.