Server migration is a massive project. It takes time, planning, and experience to get one done. Sometimes the project takes days … but seven server migrations in one night? Impossible.
“It’s kind of fun to do the impossible.”
~ Walt Disney
David Xiong, our Vice President of Technology, agrees. He loves doing things with technology that other I.T. support firms won’t even try. Things they say are impossible; that can’t be done. Such as revamping seven servers in one night.
Sounds like a plot line from the “Brave Little Tailor.” The 1938 Disney cartoon starred Mickey Mouse who brags of killing seven flies in one blow, but the townsfolk misunderstand and believe that he slew seven giants. A fable for children which came true at J – I.T. Outsource last week. Except in this case, David and his technical team did do battle with giants – and they won.
The mission, they were told, should they choose to accept it, was to reconfigure seven servers. Normally a project that would take other companies days if not weeks, they had just one night to get it done. The client, a large organization beloved by the community, serves their market seven days a week. They could not be down for a server migration during business hours.
Any other I.T. support firm would have told them they were out of luck and they would just have to do things manually while the server revisions were made. This is why you’ll visit a website, or try to place an order, and the company will tell you they are “down for maintenance.” Sorry for the inconvenience. This, to us, is not acceptable. You hire a managed service provider to solve business problems, not cause them.
So our team, reminding themselves of the company motto (“I.T. is not for the weak.”) prepared for battle. No, they didn’t stock up on energy drinks. They planned. That and a whole lot of experience is what it takes to tear down, rebuild and boot up seven servers in under 13 hours.
This wasn’t a simple operation like putting in a new inkjet printer (“Unplug the old printer, plug in the new printer, follow the install wizard steps”). This was a major operation because every single server was configured differently. They may have started out with the same setup, but a computer network is a constantly changing, dynamic environment. Each change creates a whole new branch of possibilities which could lead to any number of problems. If the server was set up using best practices for performance and redundancy in the first place. These weren’t, which is why our technical team was going to be working on the server migration all night.
Before they could tear anything down, they backed everything up, took the virtual hard drives, moved them to a safe place, turned one real server into a virtual machine, and then set up the structure of each server using our standard best practices for performance and reliability. This takes time and then there is simply the sheer volume of data that needs to be transferred out and back in. These were large files and there’s just a limit to how fast they could be moved.
As the night grew quieter and darker and the time grew later, the team kept moving, kept watching the files move, and finally in the early hours of the morning, they were ready to test. The moment of truth – time to find out if they failed or if they succeeded.
Of course, they succeeded. Mission impossible became “Jack the Giant Killer.*”
Two things: planning and experience.
Not the simple plans we make when thinking about our weekends. “We’ll take the dog to the groomer, get lunch, see a movie and then go shopping.” Extreme I.T. support planning that takes weeks. David and his team created a 17 step plan to be completed in the timeline of 14 hours. The timing had to be strict and took into account the bottleneck of data transfer. The team then reviewed the plan together every day the week before the transfer and then spent about four hours on Friday laying the groundwork for transferring those files.
The plan also included time for testing before the client had to open at 9 a.m. that morning. By 5 a.m., all the basic operations were in place and the team was able to test and troubleshoot a few things. They also triple-verified that everything was replicating to the backup before finalizing the server migration.
When the business opened, everything was in place and running better than ever, with no one the wiser except for our exhausted team who all went home for a well-deserved rest. On Monday morning, they were back in the office and ready to go again.
This would not have been possible without David Xiong’s extensive I.T. support and deep technology development experience. He’s lost track of the number of server migrations (close to 500) he’s done but says, “I do more migrations in six months than most people do in three years. I guarantee it.” He’s probably done more in the last 15 years than most people do in their entire career. When he worked for a major phone, data and cell carrier, his call center had to migrate every one of their servers to new platforms. All 338 of them. Leading a team of 10 people, David got it done in two months. That’s 5.6 server migrations per day.
You can’t pay for that kind of experience. Literally. David says that you can’t find training courses or manuals on server migrations. It’s a dirty little secret of many computer repair guys is that they are Googling step-by-step directions on how to fix technical issues. This is also why that online chat help representative is so slow. They’re looking up the answers. You can’t Google server migration instructions.
Google will tell you where to find and how to set different settings within a server, but not how to migrate it. There are too many variables from when a server was set up in the first place and from each modification and update during its lifetime. The only way to get really good at server migrations is to do a lot of them and David has done that.
There are people who have this kind of experience, but they work for I.T. support departments of major corporations; for enterprise-level organizations. Their level of expertise just isn’t available to small and medium-sized businesses. Unless you’re here in the Central Valley and you’re a client of J – I.T. Outsource.
The Problem with Doing the Impossible
Once you’ve done the impossible, you’ve got to do it again. Fortunately, our team likes that. They enjoy doing things no one else will try; things that others say can’t be done. They relish the challenge. Once again, they seem to be channeling the master of imagination:
“We keep moving forward, opening new doors, and doing new things, because we’re curious and curiosity keeps leading us down new paths.”
~ Walt Disney
If your I.T. firm or managed service provider has ever told you “No, it can’t be done” or that you’ll need to be down for more than a day, then you need a new provider. Call me now at (559) 485-4335 if you want to talk about your I.T. support.
*Also produced by Walt Disney as a “Laugh-O-Gram” in 1922.
Photos from Flickr (top to bottom) courtesy of: