As enterprise systems are evolving and moving to the cloud, I have been diving into whether or not DaaS is finally ready for enterprise adoption.
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
Is it ready? I am finding that the short answer is “it depends.” I know that you were hoping I would say yes, it’s ready, or nope, it still sucks, but the reality is that this technology has evolved greatly over the past few years. Depending on your business requirements, it could be time to move forward with DaaS.
Let’s look at what is making this option more viable and worth implementing for the enterprise.
Lifecycle management – Almost mature
I don’t know about you, but I can say that at every organization I have worked with in the past 20 years, lifecycle management has been a constantly evolving process and conversation: When should I upgrade to the next patch rollup or product version of my application? How can I ensure that I have completely patched Windows, and closed off any other system-related vulnerabilities? Really, lifecycle management is more than what’s the latest and greatest—it’s about ensuring that my organizational workloads and data are secure.
I am seeing that DaaS providers understand these lifecycle management needs. Their offerings allow you to do lifecycle management that is likely better than what your organization is currently doing onsite.
It is important to understand your DaaS provider's lifecycle management policies. Make sure that you know how they manage your upgrades—including how they will communicate that a change is occurring—so that you can notify your users.
The GPU revolution! – Almost ready
Well, it’s not exactly a revolution, but when you are researching DaaS providers, you will find that the more mature offerings will be able to provide GPUs for your desktop application graphics needs. You probably already know this, but using GPUs to offload graphics processing will improve overall application performance.
I am seeing that there are still challenges tied to implementing GPUs on-premises, due to either cost or improper planning for virtual desktop deployments, so it is wonderful to see that GPUs are available through DaaS providers.
Shifting away from endpoint support – Getting there
If you can deploy DaaS, then you can provide your organization with a workspace that can be used from anywhere there is an internet connection. This means that strategically, you no longer need to care about what the endpoint device is, and your users can use whatever they are most comfortable with. DaaS providers are empowering us to shift to a more holistic BYOD offering.
Most organizations could also get great financial benefits if they could get out of the business of managing desktops, and then reassign those existing teams for greater organizational roles.
What, I need to build my workloads on-premises first? – Needs improvement!
Many of the offerings in this space only work if you build your desktops onsite first, and to be honest I just find that for many organizations this is not realistic. While this is generally good for larger organizations, small and midsized organizations often don’t have the resources or capacity to first build their desktops on-premises and then move them to the cloud. So, for customers that do not have the luxury, DaaS providers should just let them do all their work in the cloud already, and support both options!
Service level agreements – Needs improvement!
In order to provide an enterprise-ready service level agreement (SLA), the expectation is that my systems will be online 99.999% of the time or better (planned maintenance windows not included). So why can’t cloud providers offer the same?
First let’s look at how much downtime a system would endure with a 99.99% SLA: annually the system would be down for 52 minutes and 35.7 seconds. That doesn’t really sound like much, but most enterprise applications don’t tolerate any downtime.
Next let’s look at a 99.9% SLA: we are talking up to 8 hours, 45 minutes, and 57 seconds of downtime. For any organization, that’s a lot when the service provider cannot warn you about it. If this is the SLA of an offering, then I tend to stay away.
Even if an on-premises system had an outage of this nature, the probability of it occurring during business hours would be lower. Why? Because you have control over when you apply updates and make changes in your on-premises environment. You do not have this level of control with a cloud-based offering.
In my research I have found that DaaS has developed a lot in the recent past, and that the product offerings available in this space are maturing to a level where organizations can really consider using them.
Even though there are several areas where DaaS offerings need improvement, the decision to implement DaaS can become an “assumed risk,” according to your organizational culture.
Additionally, how DaaS will be used will affect your risk level. For example, if your primary use case is for consultants or disaster recovery, then the underdeveloped aspects of DaaS may be less relevant. As for DaaS for production, I am still warming up to this idea.