The Fallacy of the Cloud

Standard

As part of my efforts to formalize much of the IT experience and education I already have, making it easier to present and qualify, I took the opportunity of the recent Christmas/New Year break to start going through Rackspace’s Cloud University (CloudU) curriculum. I actually took the time to read all of the material provided for the 10 lessons, eschewing the hour long webinars. I also passed the final exam and obtained the certificate (you can see my badge on the right).

Much of the material, concepts, and decision points discussed were very familiar to me having researched and written extensively about cloud computing during my time at Info-Tech Research Group. However, one thing kept jumping out at me about the information compiled by Ben Kepes of Diversity Limited; this thing about a supposed lack of system administrators required for those organizations choosing to deploy their infrastructure in a public cloud (i.e. Amazon AWS, Rackspace Cloud, GoGrid, Flexiant, Joyent, CACloud, etc).

Mr. Kepes and I are both members of the Cloud Computing Standards Forum on LinkedIn, and share several connections, so I took the opportunity of being a 2nd degree connection to take a look at his profile. As I suspected, Mr. Kepes does not have an IT background. As far as I can tell, his experience with computers does not go to the depth required to actually go about deploying any cloud infrastructure, whether public, private, or hybrid, and so I’m no longer surprised about his conclusions.

In reality, he’s way off the mark. He consistently talks about how IT professionals will need to adapt their skills in order to ensure they’re still relevant to a business that chooses to deploy some, or even all, of its infrastructure in a public cloud. That’s true, but he assumes that not doing so will directly result in those same IT professionals losing their jobs.

Fortunately, he’s wrong. Smart, and very insightful, but wrong. The type of scenario Mr. Kepes is actually referring to is moving to a managed services provider… the kind of company that will not only host your infrastructure, but will also perform much of the system administration for you. In reality most public cloud providers, including the ones I mentioned above, don’t offer that level of service. And if they do it’s at a premium, minimizing the cost reductions a business is expecting to achieve (rightly or wrongly).

No, unless you’re training the Receptionist to configure a fresh Linux or Windows machine to run the application (effectively making them an IT pro as well!), SysAdmins are still very much a part of the future. One “skill” that moving to a public cloud minimizes/negates is how to size and buy a server. I don’t know any IT pro that’s ever put “server buying/sizing” as a skill on their resume, but this is one big step in the application deployment process that will no longer be required, or at least minimized, as public cloud infrastructure is increasingly utilized.

So, for all it’s worth, don’t believe everything you read about moving to a public cloud. It can absolutely be very beneficial, especially for web-based apps that you still want to control for yourself (e.g. SharePoint, Cynapse, an IP PBX/phone system), but don’t expect to reduce your IT staff because of it. Your employees still need computers of some sort to connect to those cloud-hosted apps, and therefore a corporate network is still required, and someone needs to maintain that network. And your employees’ machines need maintaining and support. And those cloud/virtual servers don’t configure themselves, nor do the applications you install on them. Even companies that adopt SaaS solutions will inevitably finding themselves stuck for skills to help them integrate everything.

You will, however, get out of the server-buying business, and therefore reducing your electricity costs, and capital expenditures. That’s fine, because buying servers is not what IT professionals enjoy doing all day, anyway. There’s nothing wrong with asking your IT staff to become more strategic partners to the business, and no doubt you’ll find many eager to do so. Just don’t expect to get rid of them simply because you’re not buying any boxes for the server room.

Why I Sold Orpheum

Standard

Two years ago I founded a web hosting company, Orpheum Hosting Solutions. I set out to build a web hosting company that could compete with the likes of MediaTemple on infrastructure, and HostGator on price. As time went on I added more plans, did a wholesale move from a cloud server infrastructure to big, dedicated servers, and even switched backend client management systems to make things easier for everyone.

A few months in, I started to think, “Orpheum should be a business services company. A one-stop shop for small businesses to get their website, web-based applications, and even phone lines, or a phone system all from the same place without having to call Rogers, or worse yet Bell.” So, over the course of the next 36 months, Orpheum began to expand into managed servers, hosted PBX, SIP trunks, and eventually virtual private servers via the acquisition of AeroVPS.

Orpheum was doing well. It was not, however, doing well enough to pay anyone a full-time salary. Two years in Orpheum had acquired several high profile, lucrative customers worth over $1,000 a year. Unfortunately, due to a mix of issues with billing systems, and user interface challenges, onboarding these customers was time consuming. Too time consuming when you factor in that I still had a full-time job, and a young, growing family. Too time consuming when, through my own fault, each VoIP customer required several hours of assistance to get online.

So what happened? Orpheum expanded too quickly. I felt like I spent a lot of time finding the right solutions, and partnering with the right vendors. However I don’t feel like I spent enough time ensuring everything worked properly, that I really understood the technology behind the service and what Orpheum’s customers saw on a daily basis (with the exception of the web hosting and VPS services), and ensuring that customers were onboarded properly. I started to fix this a few months ago, but again… the things I *needed* to be doing had to come first, and the things I *wanted* to do came second, or even third or fourth. Orpheum’s customers suffered.

I brought on a team of people to assist with tech support. They provided level 1 support for the web hosting service, but they were not as familiar with the control panel I used as they were with other, more popular solutions. But it still took a lot of weight off my shoulders, and let me focus on supporting the VPS and VoIP customers. It wasn’t quite enough help though.

And then I really bunged things up when I spent several months, and a good amount of my capital, attempting to build a cloud infrastructure-as-a-service (like Amazon AWS, or Rackspace Cloud) offering. My vendors were not forthcoming enough with pertinent information, over promised, and then under delivered. I blew $1,000 to find out I couldn’t do something the way someone told me I could.

So, I made a very difficult decision to sell Orpheum Hosting Solutions. However, despite the disgruntled customers and accusations otherwise, I did not sell to the highest bidder. I had multiple bidders put offers in over five digits, but I chose to sell to a company/owner that I knew was going to leverage the existing services offered, regardless of whether they kept the brand intact or not. Those other high bidders wanted to dismantle the company, separate off the VoIP customers, perform wholesale moves off of the existing clustered hosting infrastructure onto more traditional, single-box solutions. I said “No” to those bidders.

At the end of the day, this was a lesson in growing too fast, without enough resources. You can build a business yourself, part-time, but it isn’t easy. You need razor sharp focus in the first few years… something I’ve read about time and time again, but once again I make myself learn it the hard way. I will be trying again… it’s in my blood, I can’t help it. But next time the messaging, the onboarding process, the support… everything will be polished, ready for customers on day one. And I’m going to focus hard on the core services people want and need out of the company.

SQL Server 2012 – Best New Features for Small Business

Standard

I recently read through Introducing Microsoft SQL Server 2012, the book published by Microsoft Press. It’s a rather large book, and going through it I saw a lot of terrific features for medium and large enterprises, especially from a Business Intelligence point of view. However, I also pinpointed some features that I feel will be very useful for small businesses, so I wanted to give them a bit of explanation here. This is not going to be a series, just a single blog entry.

Multisubnet Failover Cluster

SQL Server 2012 features much better failover options than SQL Server 2008. My favourite, Multisubnet Failover Cluster, simply wasn’t possible in SQL Server 2008 because a failover server had to be on the same subnet. In SQL Server 2012, failover servers can be on different subnets. That means no VPN or VLAN is required between servers in different locations or network segments, which also frees up the possibility of putting your SQL failover server in a remote datacentre, whether it’s a colocation, virtual private server, dedicated server, or even a cloud server.

Support for Windows Server Core

SQL Server 2008 had to be deployed on a full-blown Windows Server instance. That’s no longer the case. SQL Server 2012 can be deployed on Windows Server core 2008 or 2012. This frees up additional resources for SQL Server, reduces the attack surface of Windows Server (therefore providing better security), and also reduces the number of patches that need to be deployed to the server overall. It’s a big win, but be sure to look at all of the things that are not supported when deploying on Windows Server Core. The list isn’t that bad, but you should be aware of the limitations.

Database Recovery Advisor

Backing up and/or restoring a database is now a much more visual experience in SQL Server 2012, and in a good way. The Database Recovery Advisor now gives you a visual representation of what point in time you would like to take a backup from (assuming transaction logs are available), and the same interface is provided for restoring backups. So now you can easily take a snapshot of a specific point in time, and restore back to that same point. Having a great deal of experience with this procedure, I assure you it’s a massive improvement over the current method.

Audit Supported on All Versions

SQL Server 2008 introduced the Server Audit Specification, and Database Audit Specification objects. These specifications were used widely for auditing and compliance requirements, but many users were not satisfied because these features were only available in premium versions of SQL Server. Many users had to rever to using SQL Trace instead, which brought about challenges. The Server Audit Specification and Database Audit Specification objects are now included in all versions of SQL Server 2012, and SQL Trace will likely be retired in the next version of SQL Server.

Contained Database Authentication

SQL Server 2012 now brings the ability to create users authenticated for only a single database, and transferring the database to another server also brings the user setup with it. That means greater portability for databases (important when moving a database to a more powerful/its own server), easier configuration of failover clusters, and just easier administration overall.

Those are the insights I wanted to share with you regarding SQL Server 2012 in a small business. Any other enhancements have been made around Business Intelligence capabilities which, while very interesting, aren’t critical for the day-to-day operations of a small business. Plus, if you’re doing homegrown BI, chances are you either specialize in such functions, or you’re not a small business after all.

London, Ontario – A City of Bests?

Standard

Well… the past few weeks have really gotten London, and its citizens, reeling. Amongst the day-to-day stuff that happens – snarling traffic, complaints about students, store/factory closures – we also have some heartening news like progress on an HIV vaccine, and the ability to communicate with some people who are in a vegetative state. Can London turn itself around? Can we focus less on Mayor Fontana’s criminal charges and mounting legal battles, and instead focus on what the Forest City does well?

I know… in this recent era, you may be asking yourself… what on Earth does London have going for it? Plenty, actually! And I’ve come up with a brief list if things that either London is tops in, or near the top.

Let’s face it… we need to focus more on what we do well, and continue to push other institutions to do more. Let’s strive for better, take risks, continue to invest money where it makes the most sense, and prepare ourselves for the future. We cannot foresee all that will happen, especially with JoFo at the helm, but we can make the decisions necessary to ensure London survives, and thrives, even in the face of adversity at the top.

What else is London tops (or near the top) in?

Windows Server 2012 In Your Small Business – Part 8

Standard

Windows Server 2012, as always, is available in a few different flavours. The names have changed a bit from Server 2008 R2, but overall the effect is the same. Your small business will want to focus on evaluating Windows Server 2012 Essentials. Why? For a few simple reasons:

  • It’s a full-fledged version of Windows Server 2012.
  • Essentials can perform complete system backups, and bare-metal restores, of the server itself.
  • Essentials can perform complete system backups of client workstations.
  • Microsoft Online Backup, its own cloud storage service, can be used to protect data.
  • Essentials centrally manages and configures the new File History feature of Windows 8 clients. This helps users recover accidently deleted or overwritten files without IT support.
  • It can monitor the health of Windows 7, Windows 8, and even Mac OS X 10.5+ clients, notifying of any issues related to backups, low disk space, and others.
  • As nice as Exchange Server can be, you’re no longer tied to it. You no longer need to feel as though you have to leverage the investment made into Exchange, because it’s no longer included. Head for the cloud!

All this for up to 25 users, and 50 devices. Fully recognizing that employees now use multiple devices at work, the device count is higher than the employee count. Smart move.

Now, I say all this, over the past few weeks, with most of you knowing full well I’m a big open source guy. And I still am. But when dealing with smaller businesses, few are willing to entertain the idea of completely ditching Windows. So even if they employ a bring-your-own-device (BYOD) strategy and let power users bring their own Ubuntu Linux laptops, or Apple iMacs, the vast majority of businesses are still running Windows in the server room/wiring closet.

And that, dear friends, concludes my series on leveraging Windows Server 2012 in your small business. I truly hope you found it informative and worthy of the time you spent reading each entry. If you have any questions, feel free to leave a comment or send me an email through the Contact page.