| The starting point is to look at the problems | | | | a certain investment of time and skill in the |
| associated with traditional business continuity and | | | | execution, but a significant reduction all the same. Its |
| disaster recovery strategies. At enterprise, | | | | this inherently efficient portability that has got people |
| multi-national level, it is estimated that almost 80 per | | | | excited because here at last is a way to break the |
| cent of the Disaster Recovery budget is given over | | | | stranglehold of the one-to-one server relationship. |
| to the protection of just 20 per cent of the total | | | | Portability is the essential by-product of the real |
| server network, namely their most business-critical | | | | ‘virtue of virtualisation' – encapsulation. You |
| servers. It's an understandable risk management | | | | can encapsulate an operating system, software, and |
| trade-off. | | | | its data into the equivalent of a single discrete file, or |
| In an ideal world, organisations would want to protect | | | | workload; you can then transmit this workload to a |
| all server workloads, with Disaster Recovery | | | | secondary location, create a bootable backup on a |
| procedures geared to rapid recovery with a high level | | | | virtual recovery platform and get your users back |
| of data integrity – they would look to a separate | | | | online in a matter of hours, or less. Suddenly, |
| data centre or centres and maintain mirrored | | | | organisations don't have to be shackled by the |
| environments, perfectly synched and ready to go. | | | | heterogeneity of disparate servers, operating |
| But the reality of those hot standby environments, | | | | systems, applications and data sets – instead |
| and the high end server clustering and data replication | | | | they can leverage the homogeneity implicit within |
| that powers them, is that you need a one-to-one | | | | portable workloads. |
| hardware relationship and software redundancy: you | | | | They can now move and replicate complete |
| must have exactly the same server configuration at | | | | workloads as discrete, aggregated units, detaching |
| the discrete data centre as at the primary site with | | | | them from their native hardware configurations and |
| precisely the same operating system versions, | | | | directing them between physical and virtual hosts as |
| licenses and patches installed. It's a costly and a | | | | needs dictate – from physical to virtual, virtual to |
| complex undertaking that is justifiable only in relation | | | | physical, physical to physical etc. One could argue that |
| to ultra mission-critical servers; the rest of the IT | | | | for a hard-pressed IT team the cost savings are |
| estate will have to make do with more sustainable | | | | very much secondary to the agility and flexibility |
| but lower – and therefore more risky – levels | | | | afforded to them in dealing with unpredictable service |
| of protection. Think of it in insurance terms: you | | | | disruptions. That's the trouble with disaster recovery |
| arrange a reasonable level of cover at a reasonable | | | | – most people have at least got beyond the |
| price and you accept the risk/cost compromise; it's | | | | ‘it'll never happen to me' ostrich act but find |
| only when you want belt and braces cover for those | | | | themselves stymied by an endless array of |
| specified high value items that the premiums shoot | | | | ‘what could happen' scenarios and crafting a |
| into ‘how much?' territory. | | | | suitably risk-balanced response. Disaster Recovery |
| Against that backdrop it's easy to see what | | | | can come out of the mundanely routine – the |
| triggered the interest in virtualisation as a new | | | | hardware or circuit failure, the software glitch, plain |
| Disaster Recovery dimension. Even by the crudest | | | | old human error – or the spectacularly |
| measure, replicating five production servers on a | | | | devastating – the fire, the flood, the bomb, but |
| single server running multiple instances of a virtual | | | | either way you need to be adequately prepared for |
| operating system is going to reduce costs – not | | | | whether it is just one server that's affected, a single |
| a directly correlating 80% reduction and not without | | | | site or the whole corporate wide area network. |