Most of our non French users should have noticed we’ve faced a downtime that started on Dec 25th around 8PM Paris time.

This problem concerned our Moodstocks API platform, and as you probably know, since this stack powers both our clients’ applications and Moodstocks’ ones it had a noticeable impact.

In practice we faced a problem with the main server that runs Moodstocks API stack which freezed and became unreachable – even if it was still flagged in Running state by AWS. This server is an Amazon EC2 (Ireland) 64-bit EBS-boot instance, built on-top of an Alestic Ubuntu 10.10 Maverick AMI.

The Amazon health dashboard points out that some issues occurred on Amazon EC2 Ireland that precise day:

Between 11:10 PM and 5:35 AM PST the EC2 APIs in the EU-WEST-1 region experienced elevated error rates.

However Amazon says that: Running instances were not affected. So we can’t affirm that our problem is related to these EC2 EU-WEST-1 issues, even if we suspect it could be.

At that time, we did our best to restore the service in the shortest time frame and make sure no data was lost. All our indicators are OK. However, if you suspect any problem with your API keys and/or our SDK usage, feel free to contact us directly and we’ll help you investigate what’s going on.

We’re really sorry for this inconvenience. We’re going to keep working on improving all the processes that contribute to the reliability of our API platform.