Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

UK firms face more unexpected cloud problems than global counterparts

Margi Murphy | June 27, 2014
More UK enterprises experience at least one unexpected challenge when moving to public cloud than anywhere else in the world.

UK businesses are more likely to experience challenges when moving to a public cloud than firms in other countries, according to a survey.

While 88 percent of enterprises across the US, Europe and Asia Pacific said that they experience at least one unexpected challenge when moving to the cloud, the UK was higher than average with 93 percent, a study commissioned by VMware and disaster recovery as a service provider (DRaaS) Iland has found.

Of the 415 enterprises surveyed by Enterprise Management Associates (EMA), pricing and performance were the key challenges. Issues with support affected 36 percent of respondents and downtime plagued 35 percent. Management of their cloud services and scalability was the least reported challenge, but still caused problems for 33 percent of businesses.

The report states: "The most common challenge in the US was support, while performance and downtime topped EMEA and APAC lists, respectively."

Public cloud performance, Europe's biggest challenge, was affected by "noisy neighbour" syndrome, the study said, referring to when too many users are on the cloud. Moving to the cloud can even accentuate noisy neighbour issues a business may be having within datacentres.

Although the report finds that "many have fallen victim to cloud struggles", analysts have said that these challenges may be down to more than just the choice of vendor.

Failure rate by vendor

Of the medium to large enterprises surveyed across UK, Europe, North America and Asia-Pacific, the highest stall or failure rate was attributed to Rackspace and Amazon as IaaS providers.

Respondents were asked which public cloud platforms they had considered, and whether they had stalled or failed, or adopted the platform. Out of 415 enterprises who used Rackspace for their cloud, 63 percent of projects stalled or failed, 57 percent of Amazon users suffered stalls or failure and 44 percent of Microsoft Azure users experienced stalling and/or failure. The lowest failure rate of 33 percent was attributed to VMware's vCloud, the vendor that co-commissioned the study.

Quocirca analyst Clive Longbottom warned that choosing a public cloud vendor is meaningless without a robust infrastructure.

He said: "We [Quocirca] tend to find performance and scalability are two things that are inextricably linked, and that it doesn't make any different if you are Amazon or not, if your application is not architected correctly it shouldn't be in the cloud anyway.

"If you have something that is there [in the cloud] and it starts behaving like a dog, you start throwing Central Processing Units (CPU) at it. It improves by five percent but it should have improved by 50 percent, because all you have done is move the problem on to input/output (I/O).

"Then you throw more storage at it, and then you find out, actually, it was a network problem. So you throw more network at it by which times it has become even more confusing because you had no idea this was what you needed to do. The likes of Amazon and Google, unless you are a marquee customer, aren't going to sit down and help you to work this out."

 

1  2  Next Page 

Sign up for Computerworld eNewsletters.