Current status: All our ducks are in a row.

What is this site? We monitor the status of coursera.org and our class sites, and we update here whenever there are interruptions in service.

Monday, December 24, 2012

12-24-2012 - 01-01-2013

Over next two weeks, Coursera will be performing routine maintenance on archived, inactive and upcoming classes. This will involve turning off access to the site (including to new enrollments and access to the class site itself) for up to 15 minutes.

Classes which are currently ongoing will not be affected by this maintenance. We are sorry for any inconvenience caused.

Tuesday, November 13, 2012

13-11-2012

At 6.45pm, 13 Nov 2012, Coursera's database instances on Amazon AWS went through an Amazon initiated and managed system upgrade. While most of our database instances got upgraded without much issues, one of our database instance met unexpected issues and is in the process of recovering. This is affecting a small subset of our classes and our student wiki.

Sunday, October 28, 2012

10-28-2012

Starting at Sun 28 Oct 2012 7:30:00 PM PDT, Coursera will be performing preventive maintenance on our systems ahead of Hurricane Sandy. During this time, some class sites may not be available for up to 15 minutes. We apologize for any inconvenience caused.

Monday, October 22, 2012

10-22-2012

Starting around 10:30 AM, we've  experiencing issues with our hosting provider (Amazon).  See http://status.aws.amazon.com/ for details.  We'll be back up soon!

Update (11:30AM PST): AWS issues seem to be more widespread than Amazon had anticipated and affecting a lot of our disks / storage on the cloud in multiple availability zones. We are contacting and working with AWS employees to resolve the issue. We are sorry for the inconvenience caused and hope to be back up soon.

Update(1:22PM PST): We're back up! Thanks for your patience.

Update(9:15PM PST): Our databases had intermittent issues again from 8.45PM -- 9.15PM PST. We're  back up again.

Wednesday, October 3, 2012

10-3-2012

We experienced downtime from 4:45-5:00pm PST today while we were migrating away from a database system that has been causing problems over the last few weeks.

Tuesday, October 2, 2012

10-02-2012

Our users are seeing intermittent service interruptions. We apologize, and are looking into the issues.


Update: The issues have been resolved and the site should be performing fine now. We experienced some intermittent database issues and have manually transitioned to more powerful machines; we're in the process of moving off the flaky sub-system by the end of this week. We apologize again for the inconvenience, and will closely monitor our systems for the next few hours to ensure site stability.


Friday, September 28, 2012

9-28-2012

We experienced downtime on class.coursera.org and www.coursera.org from approximately 4:25 to 4:45 PST.

We traced the outage to an issue with our MongoDB database and are taking steps to prevent it from happening in the future, and to recover faster.

We apologize for the outage and have emailed affected students to let them know.

Sunday, September 23, 2012

9-23-12

We experienced downtime on class.coursera.org and www.coursera.org from approximately 5:00 to 7:00 AM PST.

One of our critical distributed databases went into an inconsistent state. When the database is in an inconsistent state, it brings down all class websites. We were already in the process of transitioning off this database system, and hope to fully decommission this system within 2 weeks. We apologize for the inconvenience this caused.

For students working on assignments that were due, we moved the deadline forward.

Monday, September 10, 2012

9-10-12


Summary:
Today, GoDaddy, our DNS provider went down in an attack, bringing coursera.org and others with it. (BBC, ZDNet)  We worked quickly to transfer all our DNS off of GoDaddy. We have now restored service using Amazon's Route 53 service, and are back online.

4:30pm PST: All services operational. Our TTL should have expired globally, and service should be back to normal.

2:15pm PST: We were able to reach GoDaddy's web interfaces, and have transferred our DNS name servers to Route53. Services should be returning as the TTL on GoDaddy's name servers expires.

11:45am PST: We are unable to reach GoDaddy's servers ourselves, and are locked out of our DNS settings.

11:30am PST: We are getting reports of DNS not resolving. Investigating...