Amazon Web Services experienced an outage in its U.S.-East data center on Christmas Eve, taking down both Netflix and Heroku. Service wasn't fully restored until Christmas Day.
It was not a merry Christmas at Amazon Web Services' U.S.-East data center. Of all the times for the data center to experience unexpected downtime, Christmas Eve is perhaps one of the worst in terms of timing. After all, the U.S.-East data center hosts Netflix, which, along with other customers, went down starting at 3:50 pm ET and was not fully operational again until Christmas Day.
For Netflix customers, the timing probably couldn't have been worse. Hopefully anyone hoping to catch A Charlie Brown Christmas on demand as they waited for Santa to fill their stockings had a backup plan.
Reports on the outage indicated Amazon's Elastic Load Balancing service in the East Coast data center facility was the cause of the problem. Issues with Elastic Load Balancing took the data center down. However, it seems clear the entire data center wasn't down. As Netflix and Heroku both suffered downtime from the U.S.-East data center issues, others with their applications and data in the facility were spared.
For instance, Netflix competitor Amazon Prime Instant Video, which also resides in the AWS U.S.-East data center, was unaffected. Those affected, though, were still experiencing problems into Christmas morning.
AWS outages aren't exactly uncommon, particularly in the Virginia-based facility. U.S.-East is the oldest and largest data center in the AWS network and has suffered a few outages in recent memory. The data center went down in June and October.
Such unfortunate timing for yet another AWS outage must be getting Amazon's account reps a bit of an earful today, and this can only look good for the public cloud provider's competitors.