| “We don’t need that, we’re not a computer company” or similar comments I hear quite a lot when it comes to backup solutions, UPS or security. At B Squared we take these seriously. All our servers, hard disk arrays, network equipment and phone system all on UPSs. We make multiple backups of everything. We have multiple recovery solutions depending on the severity of the problem. In the event of a total power cut, our servers, network and phone system will remain running for around 20 minutes before commencing a normal shutdown.
I visit schools and talk to companies who say “it will never happen to us” or “it will be alright” and they never spend time or money planning for the what if moments. A power spike or power cut, a hard disk failure or some other component failure or even a virus can render your server useless. These can happen at any time and are out of your control, yet will stop your business or school dead in its tracks. You may not be an IT company, but most companies or schools rely on computers, it might be documents, communicating or specialist software. Take away the computers and what will happen? Can you still function?
It is only when companies or schools experience a failure do they look into preventing failures or how to recover from them. Buying a UPS for your server will protect it in the event of a power cut and buying an external drive and backing up your data can keep your data safe in the event of failure. This is the bare minimum you should do and every company or school should do. Hard disks are cheap so you don’t have to choose what to backup any more, you can and should backup everything and multiple times. You can keep your data safe, but you also need to think about how to get the server back up and running after a failure or how long it will take to replace it with a new server.
Spending time investigating this and implementing solutions can save time and money in the future. How much money would it cost your company if you lost all your documents, data and emails? By backing your data up on a regular basis, you may only lose access to them for a few days and by having a planned recovery solution, you may only lose access for an hour or so. It is the same in schools, teachers spend a lot of time creating planning documents, resources, write reports, record assessments and teachers also take photos and record evidence of pupils work. The data is invaluable, and if it is lost some of it would be lost forever and the resources, documents and assessment information would take weeks, months or even years to reproduce. Below is a an example of disaster planning for dealing with hard disk failure;
|
Failure Type: Hard Disk
What is affected
| Primary hard disk: The whole computer would need reinstalling, if this is the domain controller, the whole network would go down and nothing would work. |
|
Data hard disk: All the data on the drive would be lost, documents would need to be recreated, some would be lost forever. |
Cost to business
| Primary hard disk: May lose whole computer system for a week or more while its being rebuilt. Additional time required to fine tune the network. |
|
Data hard disk: All data will be lost. Some data cannot be replaced, other data will require a large number of hours to recreate. This will also increase response times to customers due to not having the data. |
Either drive will cost a lot of time and money to the business.
Prevention
Take regular backups of the server including system state and images of the primary hard disk that can be used to rebuild the server quickly. Some data may still be lost, backup interval will affect how much.
Additional Option: Set up RAID on the server (replace server if needed), mirror both hard disks or use more advanced RAID options (RAID-5, hot spare etc.).
Recovery Plan
Without RAID Implemented
Primary Hard disk: Replace hard disk, use backup image to rebuild the server, restore latest system state backups and other backups. Timescale - 1 day to re-image server and restore backups. |
|
Data Hard disk: Replace hard disk, restore backups. Timescale - a few hours depending on the amount of data that needs to be restored. |
With RAID Implemented
No recovery needed, as redundancy would prevent failure. Would need to replace dead drive and rebuild the RAID array.
|
|
Its very simple to see that taking backups would save the business a lot of money compared to losing the hard disk and the time lost rebuilding the server and the network. Implementing RAID would prevent the failure from happening and there would be no cost to the business in terms of data or downtime.
Create a disaster recovery plan. List the different failures that may affect you, what is affected by the failure, the cost to the business, how it can be avoided, how to get up and running again and how long it will take to get back up and running. You can then decide which precautions you need to take to avoid failures, which you can afford and build your recovery solutions based on this information. |