Last month’s much publicized outage at Amazon’s cloud service has many cloud users and potential cloud users wondering about the reliability of cloud-based infrastructure and applications. Recent posts from this blog, including The Effects and Inevitability of IT Failure, have addressed this same issue. When a manufacturer puts their trust in an outside resource to provide vital services, they must have confidence that those services will be there when needed – essentially 24/7 for most purposes.
Scrutinize the SLA
When considering cloud ERP service suppliers, it’s important to ensure that the service level agreement or SLA specifies the minimum “up time” or availability for the service they provide. Keep in mind, though, that up-time and down time are measured against a 24/7/365 standard – 525,600 minutes per year.
A guarantee of 95-percent availability allows 438 hours or more than 18 days of down time. Ninety-nine-percent up time allows more than 87 hours without service – more than three and-a-half days!
Another consideration is the actual service that is guaranteed.
In Amazon’s case, it was EBS and RDS services that experienced the failure but the SLA specified EC2. So . . . legally, the agreement was not violated. In any case, the remedy is a refund of the service fee for the failure period. That’s small consolation for a company that might have been unable to take orders or ship product for hours or days.
Look Under the Hood
Manufacturers are wise to note the differences among cloud providers when evaluating reliability and risk. In their rush to the market some application providers rely on too many third-party components they don’t fully understand. A software vendor who builds their entire application on a third-party platform and hosts it on someone else’s infrastructure is opening themselves up to potential disaster.
Phil Wainwright at ZDNet.com points out that software vendors must engineer around the weaknesses of the platform and infrastructure providers. However, if the weaknesses aren’t completely understood (can they be?), this is a futile effort.
Know Thy Vendor
When the manufacturing enterprise considers placing critical services in the hands of an outside organization, whether it is simple infrastructure (platform as a service, infrastructure as a service) or full-blown software as a service (SAAS), due diligence is warranted. Know your ERP vendor. Question their back-up, fail-over and contingency plans. Scrutinize the SLA closely and watch for imprecise definitions and too-precise limitations. Look for a third-party audit such as SAS-70 Level II. The audit report will provide deep insight into the vendor’s practices and capabilities.
This makes complete business sense. After all, this approach is no different than what you should do with any other critical resource – whether it is handled in-house or contracted to an outside service provider. It is more likely that a reputable Cloud ERP provider will have more complete and functional backup, fail-over and contingency (disaster recovery) capabilities than most companies provide for themselves and for their in-house resources.
Note: The inclusion of guest posts on the Panorama website does not imply endorsement of any specific product or service. Panorama is, and always will remain, completely independent and vendor-neutral.
If you are interested in guest blogging opportunities, click to read more about our submission guidelines.