SaaS support challenges IT ops admins to shift gears
SaaS means no more app support, right? In practice, IT ops teams enter a new realm of upgrade management and capacity planning when their user base migrates to a SaaS offering.
While most IT professionals think the migration to cloud services will be their biggest headache, another difficulty is lurking just around the bend: support for cloud-based systems.
Once data migrates to a cloud provider and everything starts working, IT operations must decipher the SaaS support puzzle. The cloud has as many benefits as it does negatives. Like any other IT system, SaaS applications have updates, issues and outages that require support, but in a whole new way.
Cloud operations mean a shift in the amount of control held by the enterprise's IT support team. Infrastructure as a service cuts the operations team off from the physical server and, sometimes, VM maintenance, while platform as a service abstracts IT ops still more. With SaaS offerings, the vendor does updates, and the customer has little say in it.
Please hold
Vendors promise updates will be positive and roll out smoothly, but we know that is not always the case. With the vendor in charge of SaaS updates, the schedule might not be convenient for every customer. Expect services to become unavailable or degraded, or even see full outages, without any option for enterprise IT personnel to mitigate the downtime. Outages have been around as long as IT has, but SaaS outages cause a new level of frustration. As a cloud customer, IT ops staff are now the ones on hold waiting for status updates, handling communications between end users and management.
Cloud resource drains
SaaS support requires that IT teams focus on how users access the application.
Reliability doesn't come simply from good applications. It comes from controls in the environment that runs the applications. With SaaS and other cloud offerings, the cloud provider maintains the environment where the application runs, and enterprise IT maintains the connection to that cloud environment. SaaS support requires that IT teams focus on how users access the application.
Has moving a core application to the cloud created a drain on your bandwidth to the internet? Should IT ops start monitoring internet bandwidth from a general usage side to ensure that enough exists for the cloud-based application access? Slow internet has become a critical business limiter when core applications move to the cloud. During the cloud planning phase, investigate the benefits, cost and resource management involved to add another internet connection for redundancy. While cloud reduced some internal server monitoring requirements, it did not make capacity planning obsolete by any means.
Beyond network connectivity, look at what clients use to reach their tools. The web browser -- the most common application on a desktop -- is the typical way in which users get to SaaS features. Browsers are far from static: Security improvements and changing or disabled plug-in support make for a real client nightmare in some enterprises. Don't leave the desktop support team hanging. Monitor what hosted applications require from the client side, such as a change from Adobe Flash to HTML5, and provide applicable tests and migrations.
SaaS support doesn't introduce new problems for IT -- we've all dealt with browser plug-in support changes, internet connection issues and application upgrades. SaaS changes when these issues occur. Modern IT organizations get things done by adapting quickly, but behind the scenes, they have some notification of upgrades and changes. Testing, staff training and communications are planned out ahead of time, which dramatically lessens the disruption the changes cause to users and management. SaaS-based apps shorten the support lead time. There's also a risk that the SaaS update isn't compatible with an enterprise's setup, and there aren't viable alternatives. Prepare contingencies, and be ready to make adjustments after updates.
SaaS support requires skill from IT operations. Things that once were minor systems quirks are now critical. IT staff are in a weaker position to control changes, and the safety nets in testing and preproduction don't work as they did for software hosted and managed in-house. It's a new path for IT operations support, and the first few steps will reveal more than a few potholes along the way.
