One of the beauties and benefits of adopting SaaS is that it shifts operations to an outside vendor. And doing that can significantly free up internal IT support responsibilities. Generally that’s a good thing, but if you haven’t done your homework, it can lead to a false sense of security.
What would happen in the event of a disaster hitting your SaaS provider? And how soon would they be able to get your application up and running again after a major event? For any large company, a continuity plan is vital for any type of mission-critical application. The same should be equally true for a SaaS vendor. When selecting a SaaS vendor, it is important to drill down and fully understand the infrastructure and skills your vendor has.
Often, especially for smaller companies, vendor infrastructure, knowledge and experience can easily trump whatever IT resources the company itself could pull together on its own, and in these cases, if a disaster struck, the vendor would be better able to deal with the emergency. It is important to choose your SaaS vendor wisely.
Consider the case two well known collo companies in the San Francisco Bay area. In November a truck driver hit a transformer that powered the Rackspace datacenter that was home to many well-known Web 2.0 web sites like GigaOm and 37 Signals. After the power came back on, the cooling system then failed to start up, delaying the center from going back online, causing them to be offline for more than 5 hours. A second example occurred earlier in the year when a power surge on an in-coming PG&E line to San Franciso’s 365 Main datacenter causing a shutdown that lasted a few hours, pulling offline sites like craigslist, technorati, and livejournal.
A few hours offline may not seem like that long. But these two datacenters are considered to be best in class, and the incidents show how even when companies have disaster plans, problems in real life can cascade in directions not imagined during planning.
