As many people know we are in the age of the cloud. The entire purpose of the cloud is to minimize implementation timelines and reduce costs. Typically, with these savings come some risks.
Risks of the Cloud
- Hidden Costs
- Lack of Accountability
- Insider Threat
- More Silos
The cloud isn’t all bad, but there’s an old saying of “junk in, junk out.” If you move all your servers to the cloud, all you did was move your problems. You need to optimize your environment. You still have vulnerabilities and things to patch – you just moved them into a shared environment. A shared environment that is still prone to the same vulnerabilities as your previous environment, only now you don’t know what the underlying software and hardware is.
A one-time move can cost thousands of dollars! You have data storage costs and network bandwidth to think of. These are things you did all in-house previously that you took for granted. You’re paying for long term storage of data and the growth you accumulate every month. Integrating Apps become even more complex when you think about trying to integrate a cloud hosted email provider with a different cloud hosted UC solution. Remember the goal of moving to the cloud was to have that instant on and off capability. This complicates things a little bit. It’s not impossible, it’s just complicated!
Embedded costs such as power and rent are not always embedded into IT budgets. These are things you had previously that you took for granted. Well now all these costs are embedded into the cost of a hosting provider. By having to include things such as power costs and floor space this can negatively increase IT budgets. To be frank, these are IT costs. They should be included in all costs! If you purchase a server that consumes enough floor space to warrant the need for a huge warehouse it needs to be a factor of the purchase!
Lack of Accountability
I’m the most important customer! I can easily hold my cloud provider accountable! Well, I hate to break it to you, but that’s not true. In September 2015, Amazon had a massive outage that took Netflix, Airbnb, Tinder, and IMDB offline. It was huge! You have to realize these are high dollar spenders that 100% require their services to be online or it will have a significant impact to their revenues. There wasn’t an apology, or a viable explanation for the outage. Now think about government users where the outage costs lives or a failed mission! I’m not trying to bash Amazon, it’s just a scary truth about shared resources that you have little to no control over. I don’t know about you, but I like to be able to control my destiny.
Data breaches are now becoming every day occurrences. Dropbox, Ashley-Madison, OPM, ADP, IRS, and more. The point here is that it’s a serious threat! Your virtual machines will be no less prone to security vulnerabilities just because it’s moved to a shared service provider. You still have to patch and maintain all of you servers. The real issue here is you inherit all vulnerabilities of the software and hardware of your cloud vendor. Who is liable when a security incident occurs? What do you do? Read the fine print. Most cloud provider’s customer agreements say if your operations are down because the cloud provider has an outage, then they are not responsible.
Insider Threat has never been more of an issue than it is today. Data is available at lightning speed and the everyone keeps track of everything on everyone forever. Think finances, medical records, phone, text media, social, etc. Who has access to what information is not only a concern for the government, but by HIPAA and Sarbanes-Oxley. You don’t want anyone to have access to your private medical records, financial records, or better yet, the private conversation you were having with a spouse. This is what is at stake here. Insider threat and the cloud is where things come off the tracks. Do you know exactly who has access to your data at the cloud provider? Should they have access to it? Do they have the proper training to handle the type of data you have?
The biggest complaint I have about the cloud is it creates yet another silo for your organization. Or better yet, it can create a virtual junk drawer to throw everything in and forget about it. As a small business, we internally have seven cloud applications that each have a username and password. You have to manage all the provisioning of users, authentication, licensing, and more and more. The silos that exist today don’t just go away because you moved to the cloud. They are amplified under a magnifying glass. You now need a cloud authentication provider to manage all your user accounts. You still have a networking team, a storage team, an Active Directory Team, and a security team. These things are all compartmentalized. It’s just virtual and not in your hands.
What do I do about the Cloud?
The intent of this post was not intended for you to think the cloud is entirely risky. It’s intended to make you change your conversations about the cloud and to think it through. I personally like to control my own destiny. If you like to control risks and be the owner of your destiny, then sign up today to see if you qualify for a free data center assessment.
- Top Data Breaches of 2016 – http://www.crn.com/slide-shows/security/300081491/the-10-biggest-data-breaches-of-2016-so-far.htm
It took me a while to coin this phrase that we use internally at DH Technologies. I like this term because it describes our company well. We try to think big picture and always look to the future. This allows us to focus on emerging technologies and evangelize the ones we believe in. We are often called the loudest group in a crowd. I don’t mind being with the loud group; in fact, I encourage it. It’s ok to stand out! I think in this day and age the outcasts are the ones who become the most popular and successful.
What does it mean to be strategically rowdy?
You have to think about things differently. You have to look at things in their most basic form and try to get a better understanding. Question why it is the way it is. I was taught to not blindly follow rules but to get a better understanding of the purpose of the rule. Unfortunately, we all get hung up on hearsay and bad information, and it creates a bias that is difficult to overcome. You most certainly CAN NOT do business as usual.
You have to anticipate the future! Think critically! Rebuke biases and current beliefs; both your own and your customers. Interpret the information at your disposal and look for opposing views. Get someone to challenge your way of thinking so you can see different perspectives. Educate yourself so you can be prepared. Then go back and educate yourself again and again! All of these things combined are an unorthodox way of thinking. It’s ok to be different. Finally, you have to take a stance! This is where the rowdiness comes in. You can’t stand short in a tall crowd. You can’t be silent in a group of people. You can’t watch as the pitch goes sailing past you. Be noticed! Be passionate! Be a disruptor! Get the word out! Be rowdy! Just don’t be rude!
We carefully orchestrated moves in the first couple years of business as DH Technologies. We started with the right technologies. We then focused on ensuring we have the right government contract vehicles for our customers. We keep looking toward the future. Maybe you can challenge our way of thinking or educate us on the next great thing. Either way we are proud of being rowdy, and encouraging creativity. All these attributes have contributed to our growth and success.
Thinking like this has earned us a few accolades such as CRN’s #1 spot on Fast Growth 150, CRN’s 2016 Solution Provider 500, and Washington Business Journal’s 2016 Top Small Technology Companies list. Obviously that’s just the begining to great things to come. We are hiring! If this sounds like something you want to be a part of apply for a position. You can be part of our creativity! Contact Us
Many organizations are moving to virtual desktops for a variety of reasons. I have had the luxury to observe both successful and failed VDI projects. For the first couple of years when we started our company, we made a majority of our money by saving failing VDI projects. Over the years, I began to think: “What do all the successful VDI projects have in common?” Well, here’s the list of things I came up with:
1. Buy In From the Top
You can’t force a new technology on users without buy in from upper leadership. Ideally, you have already aligned organizational goals to capabilities and features of VDI. I recommend a Requirements Traceability Matrix (RTM) to ensure all the requirements are met, but that’s for another post. Upper management and leadership needs to be onboard with the changes that VDI introduces to any organization. If upper leadership doesn’t believe in your mission and project goals, then what makes you think the users will? If you are wondering why you need to care about what your users think, then skip to point #2.
2. Communication, Communication, Communication
You have to be truly transparent with your users and leadership about what your plan is and how it will benefit your organization. There are countless benefits that virtual desktops provide your users, but if you can’t very easily articulate them to you will have a rocky project.
Provide pamphlets, computer based training, and user outreach for the end users. If you show users the benefits of something simple such as session persistence, which provides you the ability to move from device to device without needing to login/logout, you will immediately win over a vast majority of the users. One of the most successful outreach events we lead was done in a government cafeteria. I’ve always been a fan of lunch and learns. We had a line of twenty or more people and it was quickly begging to get longer and longer. Not only did this provide outreach to educate the end users but it got our government customer extra funding for his project.
What doesn’t VDI touch? No seriously, what does it not impact? VDI changes the user devices, network, data center footprint, energy usage (reduces), applications, licensing, management, troubleshooting, provisioning, and more. This is just another case of why you need to communicate with your users and all the other departments.
3. Pick the best and most simple technology that’s highly scalable
All successful VDI technologies share common attributes: scalable and simple. Don’t use 10 different technologies when five will do the job. Don’t use five technologies when three will do just fine. You really have to keep it simple. Why, you might ask? Well, if I have ten technologies, I have users and administrators trained in ten different things. I also have the potential for ten different things to fail at some point which increases my trouble tickets for the help desk. Not to mention the decision tree for troubleshooting for the help desk is complex and long which increases the time it takes to close a ticket. I’m not saying this is always true, but generally speaking it is. We have been leading our VDI deployments with hyper-converged solutions which take out the complexity of deploying VDI. How? We eliminate the installation time because the hyper-converged solutions we deploy are deployed in an automated way which cut install times down to hours and not days or weeks. Additionally, I don’t have to have a SAN admin, or someone to do zoning or masking. Not to mention it’s highly scalable and predictable which makes it easy to size for small, medium, and large enterprise deployments. Ask for customer references before you chose a technology! You can thank me later.
4. Choose the right integrator
This is where things get tricky. There are two different perspectives on this and I have a biased position which needs to be addressed. I am an integrator. Lets get that out of the way. Obviously I would prefer you to use our services. You may prefer to do the work yourself for financial reasons, or political reasons. Let me explain why you should consider using an integrator for at least some of the work:
There’s a trade off between user experience and technology that is a work of art. You need to always think about the user experience in order to have a successful implementation. It’s not something you typically think about when deploying a new server. It’s something you need to constantly think about when moving to VDI. How will this impact the users’ experience? You need to communicate changes to users regularly and always before it happens.
User experience is derived from look and feel, AND Performance.
- user experience is different from Windows 7 to Windows 10
- user experience changes between Office 2013 and Office 365
- user experience is different from 1 CPU at 1GHZ and 2 CPUs at 2GHZ
- user experience is different from 2GB RAM and 4GB RAM
- user experience is different from a software GPU vs a virtual GPU
- user experience is impacted by login times, printing times
I can install this myself, I have done virtualization before
On a recent VDI project I was able to determine within five minutes that a WAN link was insufficiently sized which would cause a problem. There are things that someone with experience can quickly pickup. It’s the tell tale indicators. Lets be clear, i’m certain you can install the hypervisor stack if you or your team have done it before. After all, i just said we chose a simple and automated solution by leveraging hyper-converged solution. As someone who has overseen more than a hundred virtual desktop solutions, I can say with confidence there are many differences between a server virtualization project and a desktop virtualization project. Users will see everything you do and if things run a little slow for even a short period throughout the day, you will get several help desk calls/tickets. You typically don’t have user profiles on servers. User profiles are a majority of the number of calls/ tickets. You don’t want a good solution for user persona management, you want the best!
So what’s the solution then?
Any good consultant can help build a plan and leverage your team’s abilities without a crazy bill. There are several good approaches to building out the environment. Leverage an integrator to do a full turn-key deployment or take a hybrid approach and leverage your team with a consultant to build the environment out. Either way, do not try to do VDI without someone who has done it before. You will make mistakes. It’s inevitable, and a failed VDI pilot is the quickest way to kill any hope to deploy VDI for your organization. Besides, a subject matter expert can not only help the project out, but you can benefit from on the job training your team will receive. Think of it like a safety net for your architecture and deployment plan.
The design difference
A server virtualization project is designed from the data center out to the edge. A properly designed desktop virtualization project is designed from the user to the data center. You start with use cases and performing a desktop virtualization assessment while working towards the data center. This will help you size out the environment and determine if network segments are sized appropriately, determine application requirements, etc. Not one successful VDI project in the past five years has ever been done for more than 500 users without performing an assessment.
5. Change Management
Yikes!!! Seriously, this is more important than you think. I have been to countless customer environments where I was told, “The system runs slow, fix it.” You can’t make a little change in VDI with out it having a huge impact. For example:
- Windows Patch – No one can log in any more (This happened during patch Tuesday)
- Application Patch – all users that use print to PDF don’t work any more (Not a VDI problem)
- Network Update – a simple update moving users on a segment to an MPLS network causes the MTU size to drop by just a little bit and now everyone on that network segment can’t connect (this happened to a customer!).
- Recomposed the desktop = BOOT STORM (lets be honest, that was a 5yr ago problem)
Don’t fret – VDI can ease a lot of these issues. If you mess up a desktop image, you can simply revert back to a previous version of the snapshot and push all the new users trying to log on to the previous image. Same for applications: you can push the previous version back to users. You’re kind of hosed on the network update if you don’t have any easy methods to undo that change.
You saw the part where VDI touches everything, so make sure VDI is a priority during other enterprise changes so that something that appears to be simple doesn’t have an unintended impact to your VDI deployment.
Get a test system
You should definitely get a test system to perform updates and test patches on. If you choose the right technology, you can get a scaled down version of what you deployed to do all your tests on. The goal is to reduce outages and trouble tickets without increasing the bill.
VDI can make your end users and organization more agile to meet new demands and keep up with the ever changing world that we live in. It’s a lot harder to implement these new capabilities and features without upper management buying in on the project. By ensuring communication is constant, it will ensure you have happy users and happy administrators. Taking the time to evaluate and choose the right technology will not only make the difference in success, but it will have an impact on training and administration. Choosing an integrator with hands-on experience brings subject matter experience that is vital to speeding up a deployment and making it a success. Change management isn’t sexy, but it means that you have a great technology that stays running long-term with little to no service interruptions.
See related posts:
Should my desktop team or my server team own VDI?
How to optimize your virtual desktop image and save up to 40% resources and increase performance