Program with projects that support it. I have used this approach for longer than I care to admit in public,…
An HIT Moment with ... is a quick interview with someone we find interesting. Andy Hoover is IT director at WoundVision, an Indianapolis vendor of risk assessment software and thermal imaging tools for early pressure ulcer detection. The company recently migrated its platform from Amazon’s EC2 public cloud computing to a virtual data center.
What led you to originally choose cloud-based hosting instead of self-hosting for your application?
We are a small startup company, less than 20 employees. With limited financial resources and a small staff, there really wasn’t another option for us. We needed to be able to focus all of our attention on developing our line of products and rely on a vendor for providing a reliable hosting platform.
What did you learn about the differences among cloud computing providers?
Nearly two years ago we first looked at the big two in Amazon and Microsoft. Microsoft wouldn’t work for us because of limited capabilities with SQL Azure compared to SQL Server and the inability to install third-party software and tools on Windows Azure.
Amazon allowed us to run a little more like a traditional data center. We actually ran in the Amazon cloud for a year. But due to lack of readily available support, the learning curve of using the EC2 and S3 storage both from a development and administration standpoint, and limited monitoring and visibility options, we decided to look in another direction.
Once we decided we needed to check into other vendors, we looked at multiple vendors a little closer to home. The big thing we discovered is there are a lot of companies jumping into the cloud hosting business.
The key factors to us in selecting a new vendor ended up being the experience in the market, support options, and the physical data center itself. The provider we picked excelled in all of those areas. Bluelock has been around since 2006. Their support options and capabilities were far beyond what others could offer. There has been nothing we have asked for that they couldn’t provide or at least offer a contact for. Their data center is extremely impressive.
What special needs did you discover you needed to address because you are dealing with a healthcare application and hospital customers?
The question of "where is the data hosted?" always comes up. It became very import to be able to answer specific questions about where the data was hosted and how our data is being protected. To be able to explain where the data center is at and exactly how it operates was very important. It helps boost our credibility when talking with clients about protecting their data.
Having readily available documents from our provider, such as a SAS70 certification or a disaster recovery plan which could be passed on to clients, is very helpful. With logging being so key in guarding medical data, we found we need to be able to gain visibility into all incoming and outgoing traffic.
What advice would you have to a startup considering EC2?
While cheaper than many other providers, EC2 will require more personnel time to build and maintain. When you have questions, you are left to figure them out for yourself via knowledge base articles or blogs. As a startup, it might make sense to pay a little more to work for a vendor that functions a more like a traditional data center, has better support options, and knows who you are as a customer.
What resources did you need to implement your current cloud solution and what’s involved with maintaining it?
We needed a highly available platform capable of running multiple Windows VMs, multiple VLANs, SQL Server, and a firewall in which we have visibility.
Now that we have been up and running at Bluelock for over a year now, not a lot of maintenance is required on our part. I use their monitoring portal to keep an eye on things such as performance, availability, and usage. We are able to ask for custom options, such as custom monitoring and alerts for metrics we care more about. Maintenance of the servers doesn’t included much on our part — monitoring, patching, and pushing new releases of our software.