Disaster Recovery Exercises with the Cloud…Do it

You may have heard that SalesForce was down for around 18 hours for many of their users in mid May 2016. This outage got me thinking, what would happen to most customers if their customer relationship management software was down for a good amount of time?

How much would it cost in current sales, future sales, customer confidence?

How easy is it to get a hold of someone and how accountable is the person you want support from?

What’s your disaster recovery plan?

With the cloud we are typically giving up some measure of control in exchange for more predictable costs, ease of accessibility and simply because of the “everyone else is doing it” factor. You still need to understand though that even though you are no longer using your own hardware, your data is really what matters. It’s your data that makes up the spirit, feasibility and continuation of your business. Just because you’ve moved to an online productivity service such as Office 365 or Azure, you still need to understand that your data security, disaster recovery and access rights to your data is still your responsibility and as you are moving to a cloud based service, you should be completing a disaster recovery plan.

If I were designing a disaster recovery plan it’d roughly look something like this.

Disaster Recovery Planning (Answer these questions)

  1. What happens to my company if my email/customer relationship management etc were down for an extended period of time?
  2. Can I maintain some kind of on premise system to bring up in case, so we can keep some of our business running? Can that data be easily sycned back to online should the system come back up?
  3. What’s missing?
  4. Do/Will I still have the right staff in place to ensure that this happens?
  5. Does everyone have the proper training on the new system?

There are common misconceptions about the cloud that we can depend on Microsoft for stability and security and just forget the rest. I would argue that yes Microsoft does provide an extremely reliable environment, they system is extremely complex and there are outages that will affect your company. In many instances these outages are more frequent than your on-premise environment, but they are rarely talked about during the sales phase.

Know your SLA’s and know what they aren’t

SLA stands for service level agreement and they are just that. They are not a guarantee of up time or data accessibility. They are a standard set where you can demand some compensation for a service that was not available to you. In many cases and extended downtime might have cost your company more than what you will get back from Microsoft, you should know and understand those numbers. Typically SLA’s from Microsoft looks something like this

If the service is up less than 99.9% BUT more than 99%, you would get a credit of 25%

If the service is up less than 99% BUT more than 95% you would get a credit of 50%

Any less than than 95% and your credit is 100%.

Other things to remember…

  • SLA’s are measured on the month
  • The calculation to look at uptime looks something like this User Minutes – Downtime / User Minutes x 100 = uptime percentage
  • SLA’s can be different for Microsoft products.
  • You, the customer must initiate the request for compensation

A good guide for SLA’s can be found here¬†http://www.microsoftvolumelicensing.com/DocumentSearch.aspx?Mode=3&DocumentTypeId=37


  1. Excellent site. Plenty of useful information here. I am sending it to a few friends ans also sharing in delicious. And certainly, thank you for your effort!

  2. Very good article post. Great.

Leave a comment

Your email address will not be published.