IT companies keen to attract new business to their Cloud offerings need to work hard at building customer confidence in their ability to deliver, in actually delivering, and in responding rapidly and effectively when things — inevitably — break.
Although neither a panacea nor universally offered, the Service Level Agreement (SLA) is a routinely applied tactic that is seen to increase customer confidence, to drive internal targets, and to offer some (usually small) recompense in the event that it is invoked. Notable hold-outs are, of course, most of Google‘s services, Salesforce.com, and several other significant players in the Cloud Computing space. Even companies such as Amazon, which do offer SLAs, have done so only relatively recently (currently targeting 99.95% uptime for EC2 and 99.9% for S3). There are various reasons why the notion of a generic Service Level may not always apply at Internet scale.
3Tera yesterday announced a new SLA for their AppLogic Virtual Private Datacentre (VPDC), which will take effect from April for US-based customers. The SLA sets an ambitious target of 99.999% availability (it does not, of course, promise 99.999% availability, but nor could it), and is unusual in that it should apply credits for downtime automatically rather than requiring customers to document and claim for any outage. 0.001% downtime translates — loosely — into AppLogic being unavailable to any given customer for less than one hour per year.
Any loss of mission-critical infrastructure, even if only for a few minutes, can be hugely damaging in certain sectors. The relatively trivial recompense offered by most Cloud SLAs is actually not worth very much, and will rarely come anywhere near covering the potential loss of revenue and credibility to those relying upon the infrastructure of 3Tera and its competitors in this increasingly competitive market. In many ways, though, the subscription credits offered by the SLA are not the point. Rather, the SLA is part of the process that a company such as 3Tera must be seen to go through in demonstrating their faith in their own core infrastructure, and their contingency planning for the inevitable occasions on which things break.
Related articles by Zemanta
- Latest Google Apps outage brings down Gmail (for some) (news.cnet.com)
- Google Gmail outage compensation: $2.05 per user (news.cnet.com)
- How Google Apps Defines Downtimes (blogoscoped.com)
- Burned By Gmail Outage? Google Will (Almost) Buy You a Postage Stamp (gigaom.com)
- Can Amazon Run The Table On Cloud Services? (avc.com)
- SXSW Cloud Computing Panel: Clouds Still Need Work (gigaom.com)
- Future of Cloud Computing: A Long-Term Forecast (portfolio.com)
- Google’s Apps SLA Allows It To Minimize Downtime Of Gmail, Calendar And More (techcrunch.com)
- Storage Rages (arnoldit.com)