Report shows majority of cloud providers not hitting SLA targets


76% of organisations surveyed by CDW in its Cloud 401 report say at least one cloud vendor failed to meet their SLAs.

The report, released in February, found reliability (43%) as the most important metric for cloud service providers, far ahead of cost (28%) and ability to integrate with existing infrastructure (27%), meaning a double whammy for providers who couldn’t meet their uptime figures.

Cloud services go down for a variety of reasons, from the preventable, such as a fat finger, to the less preventable, like adverse weather conditions. Very few cloud service providers have spotless records in this department, as data from benchmarking provider CloudHarmony found.

Yet if you claim a certain figure and can’t match up to it, as Mimecast famously did, customers won’t be happy. Nor, as Verizon found out, will they be happy if you propose ‘planned’ downtime to improve infrastructure in order to reduce the risks of further outages.

CDW also found a discord from respondents between the hype cloud promises and the reality it delivers. When asked if greater flexibility in IT was a reality, 19% agreed and 7% disagreed. For ease of use, 16% agreed and 7% disagreed, while it was equal for cost savings (17% agree, 17% disagree) and a majority disagreed regarding security (11% agree, 21% disagree).

Despite this, the research found 35% of IT services today are delivered totally or partially by cloud. Going forward, organisations are considering delivering 35% of entirely new IT services via the cloud. Barriers for further adoption according to survey respondents are security (47%) – surprise, surprise – trust in solutions (31%), budget (24%) and management support (19%).

Yet the overall consensus from the report was that cloud “works”, and organisations, particularly in more regulated industries, should face up to their fears.

A CIO in federal government told the report, “Be very optimistic and open-minded about how it can enhance the efficiency and reliability of an entity’s operations”, while a healthcare CTO said, “Although it may seem complex, cloud is by far the way to go as far as infrastructure and storage.”

Do you agree with the report’s findings?

Related Stories

Leave a comment


This will only be used to quickly provide signup information and will not allow us to post to your account or appear on your timeline.

9 Apr 2015, 11:39 a.m.

I think that we should ask two more questions: 1) how often our on-premise infrastructure fails to meet our own SLA? 2) Have we deployed our application properly in a cloud? I think the first answer will be "No." for that reason we are looking a cloud providers with best SLA. And in my humple opinion, in most cases, second one will also be "No.". There is many availability mechanisms in cloud, that we are not using because of something (in most cases it's $ savings). In example in Azure Cloud, where I'm doing most of deployments, there is something like Availability Set - logical mechanism, that is immuning app for downtime caused by failure or update. Many of my client do not want to use it because of costs - they must pay for n+1 VM's to reach SLA and high availability. After failure they are blaming provider - but the truth is, that their application was not deployed as real cloud-enabled one. They use cloud as VPS hosting...