There’s no denying the importance of backup and disaster recovery plans. The goal of any DR plan is to keep data secure and keep downtime to a tolerable minimum. With many spinning plates, it’s easy for MSPs to forget these four important essentials that will ensure success when their plans are tested for real:
When you tell clients how expensive downtime can be, they’ll begin to wonder what it takes to reduce it. This leads us to recovery objectives. Without recovery point and recovery time objectives, how can you know what a client’s tolerance for downtime is? How do you know what specifications your DR plan must meet? First, set recovery objectives with your client. These usually hinge on their downtime tolerance and budget. Without goals, it’s difficult to say whether a DR plan is ever really successful. Next, build your DR plan to suit.
There’s a long list of items to include in a DR plan, but as you know, redundancy is key. Most DR plans include redundancy for things like backups, power, and communications. But to really fortify a business against downtime, more redundancy doesn’t hurt. As you look at which systems are most critical (i.e., which systems must run immediately after a downtime event), consider what redundancies to include, whether it’s for power, recovery methods, or what have you.
For some MSPs, a DR plan is a general set of instructions that never end up on paper. To ensure a quick, effective response to various downtime events, write everything down. Be sure to store hard copies in safe places and soft copies in the cloud where you can access them remotely. In general, a good DR plan will include the who, what, where, and when for various downtime events:
- What types of downtime events might you need to recover from?
- Who are the key people? How do you contact them?
- Who else might you need to contact (utilities, service provides, etc.)?
- Which machines must be restored first?
The list goes on and on. Remember that your documentation can determine your success. It can be the difference between meeting your SLAs and your client losing thousands to downtime and data loss.
Tests and Simulations
So, you created your plan. Everyone, from your team to your client, understands it and knows how to proceed if something happens. You’re ready, right? Not quite. No DR plan is complete without testing. From the individual backups to the redundancies that will keep your clients running, it’s wise to test everything. There are a few ways to approach this.
A checklist is a basic way to ensure that you can recover individual machines, and that backup power and utilities all work. This gives you some degree of certainty that you can recover, but to be positive, you should simulate downtime events. These events can be anything from DDoS attacks to hardware failure, or even power outages and natural disasters. Work with your client to schedule some time afterhours or on weekends to simulate a few different scenarios.
As you create DR plans, be certain that your client’s uptime expectations are mapped to recovery objectives, SLAs, and the equipment it takes to succeed. If you take your time developing a strategy, documenting a plan, and testing it, you’ll be ready when the time comes.