| Few companies would argue about the value of a | | | | hardware is rebuilt is unlikely to have the FD jumping |
| comprehensive Disaster Recovery plan that covers all | | | | with joy. |
| areas of the business and holds the key to | | | | For critical applications think about some 'on-site' data |
| successfully resuming day to day business activity | | | | replication. What this means in simple terms is that |
| should the worst happen. | | | | the data on Server A is replicated real time to Server |
| Most businesses would be pretty unlucky to suffer | | | | B. Should server A fail, it's a simple matter to failover |
| from major downtime due to things like fire, flood or | | | | to Server B, normally within a couple of minutes. |
| theft. Terrorism generates a huge amount of column | | | | Because the data has been replicated up to point of |
| inches and the effects of something like 911 are truly | | | | failure the users won't lose lots of data and the |
| devastating however even in the current climate | | | | system will be up and running much quicker. |
| these occurrences are thankfully few and for | | | | Of course, this provides local high availability which |
| between. | | | | although gives protection against server failure it |
| What is more likely to happen is an email sever | | | | doesn't provide any real Disaster Recovery, if the |
| failure, a corrupt database or the network being | | | | office burns to the ground, the data will be lost. |
| compromised by a virus. Guarding against this type of | | | | However, having invested in this local high availability |
| outage should be the bare minimum a company | | | | solution it's a relatively simple process to replicate the |
| should cater for, even though most of us could cope | | | | data off-site as well and deliver a true local and |
| for a few hours without email, for some businesses | | | | remote high availability and Disaster Recovery |
| this would lead to a huge loss in revenue. | | | | solution. |
| If a server failed completely, most IT Departments | | | | Disaster Recovery can be seen as an expensive |
| wouldn't promise delivery of the service back up and | | | | luxury but it really depends on how you view it and |
| running normally in anything less that a day because | | | | more importantly how you implement it. In the |
| this would mean relying on tape backups to rebuild | | | | current economic climate you can probably think of |
| the data held by the server. Commonly accepted | | | | lots of other things to spend your limited budget on, |
| logic is that tape isn't all that reliable (a side issue is | | | | however can you really afford for your main business |
| that most companies don't perform regular tape | | | | systems to be off line? The good news is that you |
| restores, so don't know how good the data on the | | | | identify the key processes that make up your |
| tape is - even if they can get is back). | | | | business and the IT platforms that support it, you |
| If a company can 'get by' for a couple of days | | | | have your starting point and like all things, it doesn't |
| without the server in place then this tells us how | | | | have to cost the earth. |
| critical to the business this particular server or | | | | One last thought for those who see Disaster |
| application is. For these servers using tape isn't | | | | Recovery a bit like insurance, do any of you regard |
| probably too much of a problem, but for other more | | | | house insurance as unnecessary? |
| mission critical application, hanging around while the | | | | |