In light of the recent outage for AWS, people are questioning the reliability of cloud services. The more appropriate question should be “is it better or worse than on-premise services?” Here are some perspectives to consider:
1. CFO’s perspective: Not My Money.
SLA’s are backed by monetary compensation. If you were smart, you put it into the contract that if you are down, they pay you for the lost revenue. You can’t get that kind of insurance for self-hosted services. The latest events are proof that no one is immune to an outage. It’s all about what happens when there is one.
2. CEO’s perspective: Not My Problem.
Their outage is not a fault of your leadership. When there is an outage for an on-premise service, customers have only you to blame.
3. Head of Marketing: Thanks for the Free Press!
If you’re a big customer of that cloud service, the press mentions your company in the news. And you aren’t the bad guy, you’re the victim, so it isn’t bad press for you.
4. CTO’s perspective: Not My People.
While the CTO might take some heat, it is much less than if it were his own people on the line.
5. I.T. administrator’s perspective: Not ME!
OK, you may already not have a job thanks to your services moving to the cloud. But if you still do, now it isn’t you having to work on this problem at 1:00 AM on a Saturday night.
Keep in mind, with a storm that has this broad of an impact, Amazon is not the only one experiencing an outage. If your company has its datacenter in the same region, you have the same risk and unless you’ve done a better job at a business continuity and disaster recovery (BCDR) plan, you’re probably going to have an outage too.
The thing about a BCDR plan is that it must be periodically tested and lessons are learned from every test. Shame on Amazon Web Services for not doing a better job of testing, but even Amazon can find a good side to this event: Now they know where the changes need to be made to improve service.