One of the things touted most about cloud computing services is reliability. After all, they are managed by the best IT people in the business running the most state-of-the art equipment.
Yet, we still hear about total system failures, and by and large, it's really hard to get a meaningful service level agreement (SLA) out of a cloud computing provider or even a software-as-a-service (SaaS) provider. Even Salesforce.com, which pioneered much of the cloud computing concept, is shy about SLAs.
In fact, even the vaunted Amazon service recently experienced an outage. According to Apparent Networks, which makes a tool that allows IT organizations to track the performance of SaaS and cloud computing providers, Amazon's Elastic Compute Service Cloud (EC2) was down for 44 minutes starting at 3:34 a.m. EST on Dec. 9th.
That may not seem like much. But according to Jim Melvin, Apparent Networks' chief marketing officer, Apparent Networks, through its service, is seeing a wide variety of performance degradation across the cloud. While full-blown outages might be rare, he notes, performance is far from consistent.
This raises questions about just how much better the IT folks running cloud computing services are than the internal IT department. It may that the only real difference is the amount of hardware they have to throw at any given application. If they really are all that much better, why aren't enforceable SLAs littering the cloud computing landscape?
Whatever the reason, it's pretty clear that customers are going to start demanding a lot more visibility into cloud computing services. And once that starts to happen, maybe we'll all see that the people building and running these IT services are just like anybody else.