Work through these steps when you first engage with a customer over a backup plan.
Audit the existing infrastructure
Look for the number of devices, the applications and the amount of data currently stored. Don’t forget to ask about laptop users, mobile devices and how many users work from home. Pay attention to the number of different operating systems and their versions as this will impact on any software or utilities you use later.
Identify business applications
Find the business critical applications and you can start to suggest remedial action that can be implemented quickly. Other applications also need to be recorded. This is an opportunity to draw up a list of different plug-ins that you will need to deal with applications.
Identify any existing backup
Look at how backup is done and identify the areas that are being missed. Pay attention to the hardware and its backup capacity and throughput. As you begin to develop the backup schedule, this will tell you whether the hardware needs to be replaced. Audit the backup logs, both system logs and backup software logs, to find errors and see how they are currently dealt with.
Estimate the total amount of data
Gigabytes or terabytes? As you audit the infrastructure you will be able to see how much disk space is on servers, storage arrays and clients and how often the contents change. Use this as a sizing guide for the different streams of backup – client, mobile, server and applications.
Read the BC and DR plans
The customer should have business continuity and disaster recovery planning documents that identify what they need in order to keep in business. This will give you priorities for the backup system.
Do a sample restore
Until you do a restore you don’t have a backup. Agree with the customer to take a random backup from the recent past – no more than one month old – and attempt a restore. Success or failure will give both you and the client a base point from which to proceed.
Build sample schedules
Focus on the different types of users and devices that need backing up. Avoid creating too many potential schedules and the system will quickly become unsupportable. Over time schedules will need to be tuned, so keep it simple to start with. For applications, use snapshots and base the time schedule on how critical the data is. This might mean hourly or as seldom as twice a day. Servers should be backed up daily using block based solutions that only back up changed data. Storage systems should be set to continuous replications. Image client machines monthly and then use a mix of full and daily backups for the rest of the month.
Identify potential pain points
These might be hardware-driven, such as lack of backup capacity. It might be a lot of data duplication that needs to be dealt with. Another pain point will be educating the users to accept the new routines.
Build backup processes
Look at what needs to be backed up and the pain points and plan the services you can offer to help the customer overcome these problems, and what you will want to charge for them.
Do a full backup
Start with key critical systems then extend the backup to cover other server and applications. Only then look at extending out to desktops, laptops, home computers and other devices. If you try and do it as a big bang, you will fail.
Only when you have been through this entire process several times will you have a complete working backup environment for the customer that can be replicated reliably.
Remember that a backup system is not something that can be built once and then left to run perfectly every time. New applications, new server deployment, virtualisation, hardware updates, software patches – all these can destabilise your backup solution. An important part of any backup management process is a regular audit of what is happening. Make this part of your service so that processes can be adapted and failures identified and rectified before data is lost, keeping the system fit for purpose and the customer happy.
Click Here to go to the main feature - Backing up, save the data, save the day