Dear Mr. Service Provider, who do you see as your biggest competitor in the Cloud market? Is it the perennial favorite Amazon with its Web services? Or the enviable Google with its PaaS and convenient apps suite?
That was the question asked by ScienceLogic in a recent survey of hosting provider attendees at the annual HostingCon event recently held in Boston. Given the amount of attention generated by AWS, and the direct correlation between the decline in shared and dedicated hosting and the growth in AWS traffic, it is surprising that AWS was not top of the list. Instead, the answer was that Rackspace (32% of responses) and Softlayer (20% of responses) were the overwhelming favorites for the title.
So why would that be? Isn't AWS providing an API being emulated by service providers and internal cloud developers, virtually the world over?
Does it have something to do with Rackspace being a large proponent of Openstack, that promises, with its enviable developer community, to re-write the IT infrastructure fabric history books?
Maybe so, but the answer lies more in the responses from a secondary question we asked hosting providers: How do you differentiate yourself from the competition? As one would expect, management and support dominated this category, at 59% of respondents. A further 32% of respondents recognized “customer control and tools” as an equally important source of differentiation.
If there was a single clear delineation between AWS and Rackspace, both vying for cloud attention, Rackspace stands out as the pioneer in selling support services while AWS pioneered the hourly billing rate for raw storage and compute power.
Putting it into perspective, there is a growing desire from hosting providers to emulate the superior management and support services that are being used in mid-size and large enterprises. And there is the recognition that support cannot be accomplished in a superior way without a strong monitoring and management platform.
Interestingly, only 22% of respondents believed that applications were a significant differentiator for them in the hosting space. Given the amount of rhetoric heard in the APM space around the importance of application performance, or at least the user's experience of that application, it is a fascinating finding.
I'm not implying that the application is not important – in fact the fastest growing revenue opportunity in the cloud computing market today continues to revolve around SaaS, which outpaces IaaS and PaaS growth by at least 50% (from a Tier1 Research study). However, the immediate need in the market is still the ability to control and manage an increasingly complex infrastructure environment first and foremost, since that is where the underlying environment is able to offer enough hints to give us predictability in our application management.
Going back to our friends at Rackspace, the company is not content to sit back and cash in on its brand reputation for fanatical support – the same feature that all other service providers now speak to. The company has recently released its new Critical Application Service (CAS) offering – which increases the level of support to end customers. The big difference however, is that CAS is specifically intended for large enterprise consumers who are far more demanding than their SMB brethren when it comes to mission-critical applications, and may be met with severe punitive action if they breach regulatory compliance.
It's no secret then that advanced monitoring is now being offered at a significant premium over and above the engineering support staff on offer under CAS. Although the Rackspace CTO downplays the relative differentiation and criticality of monitoring as simply a must have, without differentiation, it’s an oxymoron that Rackspace has premiums in excess of 10% of its high-end managed services. Specifically, offering advanced monitoring.
It's also why The Gold Plan for its new support service revolves mostly around the ability to offer full visibility to application monitoring and reporting to managed clients. The application (read Adobe CQ, and other Microsoft apps) is innumerous times more expensive than its plain vanilla, self-service IaaS offering.
I think you get the picture: Monitoring today needs to encompass all layers of the stack to appeal to the needs of contemporary enterprise architectures. Infrastructure reliability and availability is no longer enough in the modern era. The expectation is that the applications and services matter.
Antonio Piraino is CTO of ScienceLogic.