Today we experienced a DNS issue that prevented access to:
The DNS provider for these sites and services, Zerigo, suffered a severe DOS attack. Our monitoring first detected an intermittent latency problem at 11:30 PM last night. At 2:00 AM today it became consistent. Troubleshooting determined that it was caused by slow lookups to our DNS provider. The nature of the outage was such that we could not log in to retrieve our DNS records and move them until 9:30 AM. At that time, we moved our critical domains to Route 53 hosting. The switch over was completed by 10:25 AM, but users may continue to see residual issues until the TTL of the old NS records expire. This should be no later than 10:00 AM tomorrow.
If you used the local hosts file work-around posted in one of the updates above, you will need to undo it to start using the new Route 53 DNS. If you have any questions, please email us at customer-service@fogcreek.com
Following this issue, our sysadmin team will be reviewing our DNS architecture to address this single point of failure. In addition, we will update our out-of-hours incident procedures to make sure that status updates are posted in a more timely fashion. Finally, this status blog, while hosted outside of our network, used a domain that was affected by the issue. Going forward, you can access it at fogcreekstatus.com.
If you were materially affected by this downtime, please email us at customer-service@fogcreek.com and we'll make it right.
all times in US EDT (UTC-04:00)