As showcased in IBM’s announcements today, we moved our cloud systems on to IBM’s cloud infrastructure. Since this is a very big, strategic move - and potentially of interest to others in our industry - I thought I’d share a bit of background to this story.
When we started CLOUD M back in 2011, we considered very carefully where we should host our servers and backend systems. We chose Rackspace, the market leader at the time, in part due to their much advertized “legendary support”. The systems we create keep people safe, and it is therefore imperative that these systems are reliable. A key part of that is vendor support when something goes wrong and timeframe needed to fix it.
In the early days, things went well, but our needs then were much simpler than they are today. As our business grew, the numbers of servers and backend systems we needed grew as well, both in numbers and complexity. We started to run into issues, but when we reached out to the vendor for help, the results were less than satisfactory. Escalating anything via their public Twitter account triggered swift responses, however, the failures that were occurring regularly in their underlying systems were never actually resolved. Eventually, it became obvious that we would have to move on and look for a more responsive and reliable partner.
In 2014, the hunt for a new cloud provider began. We looked closely at all the key players: Amazon (AWS), IBM Cloud, Microsoft Azure and Google Cloud Platform. We spun up servers and systems on AWS and did a deep dive there. We also did a review on the other vendors published capabilities. While each of them has a unique set of technologies and differentiation, we focused on what we wanted and what was most important to us.
No. 1, of course, was support. We wanted a supplier who would become more of a companion and partner on our journey. While I’m all for automated helpdesk systems, sometimes you really need to be talking to a quaint, old human, who can actually get something done. Secondly, I wanted to avoid proprietary systems and keep as close as possible to open systems standards. I learnt some tough lessons about being tied to another vendor’s proprietary software or platform many years ago and was not keen to repeat that mistake.
A few conversations with our local IBM Cloud guys convinced me to take a closer look at their offering. To be honest, I was sceptical at first. All the “cool kids” were on AWS, so we would have needed a really good reason to not go there. However, the more I looked at it, the more I thought it was worth giving IBM a closer look, especially as their SoftLayer data center footprint was expanding rapidly around the globe and their IBM BlueMix platform offered some unique and interesting technologies like Watson - their “cognitive computing” engine.
Over the quiet NZ summer holidays, the team and I launched the process of spinning up two complete mirrors of our environment - one in AWS and the other in IBM SoftLayer. Naturally, we had configuration queries relating to our specific needs on each system. AWS had fantastic, if somewhat overwhelming, documentation. SoftLayer was a little sparse, in comparison, but when it came to support, they delivered an unexpected, extremely positive surprise. We logged a support ticket with a question on both AWS and SoftLayer and proceeded to wait. AWS’s responded seven days later; SoftLayer responsed within a few minutes!
This pattern continued. Over the course of dozens of helpdesk tickets, none took longer than 10 minutes to get an answer and every technical issue was resolved quickly. We had, in fact, completed the mirror setup and had run extensive load balancing and quality assurance tests on SoftLayer before we’d received the first response to our AWS ticket. This answered our concerns about quality and efficiency of support.
We made the decision to move our core systems to SoftLayer and completed the migration of our production systems a few days later. It’s been most interesting to see how well everything has been performing. The infrastructure is fast, reliable and the support remains exceptional. Yes, we still have a few systems running on other cloud platforms, including AWS and Azure, where it makes sense to do so, but all our critical systems now run on IBM SoftLayer.
As an entrepreneur, I am reminded every day of the importance of human support and customer care. My challenge is to maintain this level of personal attention to detail and engagement with our own customers as we continue to grow. After all, as human beings, feeling cared for and well-supported plays a huge part in our decision making, so why wouldn’t it for our customers - or yours?