All up in the cloud

Back in October of 2009 we decided to upgrade our products from our old Engine Yard cluster to the Engine Yard AppCloud. Being a small team with a lot on our plate, it took a while to redirect our full attention towards moving things over.

On July 1st we migrated Lighthouse over to the AppCloud, immediately noticing impressive results. The setup is a lot faster and more stable than ever before. As of last Thursday, our transition has been completed with the move of Tender Support.

Our old cluster configuration served us well, but we were starting to outgrow it. Engine Yard was also redirecting a lot of their focus towards newer hosting services with superior technology. Who wouldn't want to get on board with that? The AppCloud offered a number of appealing things, most notably the ability to scale the hardware easily from our end. Also, the AppCloud will give us the ability to offer dedicated instances of Lighthouse and Tender to our customers.

We’ve had a long relationship with Engine Yard. They helped us get Lighthouse off drawing boards and make it a reality. In turn, we've been able to provide an amazing service for our users and free accounts to the Open Source community. Not a day goes by that we're not thankful for their support during our early days.

Update: Tender in the cloud

Transitioning Tender to the AppCloud on Thursday night went relatively smooth. The server is faster and more responsive than we had even hoped. A number of customers have already mentioned how good things feel across the application.

As with any large move, we hit a few snags at the start. We quickly issued fixes for a number of problems we noticed on the new server that we were unable to catch during our testing. Job processing was backing up pretty heavily, so we spun up a large number of workers to keep up with the load while we addressed the problem.

We’re still seeing an issue with email attachment processing and are working as quickly as possible to get this resolved.

A number of users mentioned bugs in our marketing site last night. Those were addressed this morning. Outside of that, all other known issues have been fixed.