CIMP project update

[Ed. note: This article is the first part of a three-part series this week from the LITS Cloud Advisory Group (CAG). The introductory article about the CIMP project can be found here.]

For the past several weeks, the project team has been working hard on a pilot program to gain migration and Amazon Web Service (AWS) experience. This effort involved preparing an AWS “landing zone,” migrating applications (in an isolated environment) and then testing AWS performance. As a reminder, the following applications are included in our pilot:

  • Peoplesoft ELM
  • OnBase
  • Shibboleth
  • Confluence (Wiki)
  • Web Hosting (emory.edu)
  • Clinical Trials

We are happy to report that the results of our pilot migrations have been very promising. There have certainly been challenges, but through it all we have learned a great deal about the process and about AWS.

Photo of new employee Caleb Boyd
Caleb Boyd

Caleb Boyd, who was responsible for migrating Confluence, described the challenge of building out the network architecture: “Since the environment was fairly bare-bones to start, configuring firewalls and security groups, building out AWS load balancers, and configuring SSL certs all had to be completed from the ground up. Since this was one of the first applications we migrated, we were blazing some new trails, but the lessons we learned were very valuable in our successive attempts.”

Eddie Feliciano, responsible for OnBase, raves about the positive outcomes of the pilot migration experience, “Being brand new to this technology, everything we experience with our ‘hands-on’ migrations allows us to learn about the technical details of deploying/administering cloud solutions. Learning to think of your application’s infrastructure as code really brings out new ideas for how we could use non-prod environments to support and respond to customer issues/requests.”

As for the performance results, initially there were a couple situations where we had some concerns. But as we learned how to appropriately optimize the AWS environments, we were able to work through any performance issues and demonstrate response times similar to, or exceeding, the on-premise deployments. Caleb’s report of Confluence performance nicely sums up the results that we have seen across the board, “Our initial tests indicate that the application should perform just as well, if not better, in AWS. And even if we find that the results are lacking, the move to AWS should let us scale up extra resources quickly to compensate.”


Comments

Leave a Reply

Your email address will not be published. Required fields are marked *