On Dec. 20, I was scheduled to meet with representatives from Convoq about their new product, SellASAP. The meeting didnt take place. Why? Because Dec. 20 was also the day of Salesforce.coms outage, and SellASAPs demo accounts ran on Salesforce.coms affected systems.
While many direct users of the Salesforce.com service experienced considerable inconvenience that day, the ecosystem that Salesforce.com wants to build around its software-as-a-service platform also took a hit. You can use SellASAP without Salesforce.com, but not to its fullest potential.
So, the outage was not only a pain for users but a missed (or delayed, anyway) opportunity for Convoq.
As Salesforce.coms service becomes a platform for hosting other applications, and as customers integrate existing applications and services with Salesforce.coms sales and support applications, the risk of an outage affecting a broader audience grows.
Unless Salesforce.com finds a way to improve its own uptime, enterprise administrators wont be able to guarantee greater than 97 percent uptime for applications that touch the service.
And the more Salesforce.com partners a customer engages, the higher the degree of downtime risk: 97 percent uptime with Salesforce.com and 97 percent uptime with just one partner mean a potential 6 percent downtime window.
Services bundling isnt always a quick or easy path to lowering costs, as Sprints recent West Coast services outage proved. Software-as-a-service vendors can potentially deliver lower costs and faster deployment than internal IT departments, but weve seen that these vendors cant promise it (or at least they cant always deliver on those promises).
IT departments, meanwhile, can deliver transparency of process and established resolution procedures—and are, arguably, more accountable in the process.
eWEEK Labs Technical Analyst Michael Caton can be reached at michael_caton@ziffdavis.com.