Open Tracking has been brought back online and the issue resolved. Opens were not tracked from 8:30 AM EST to 9:00 AM EST. We’re sorry for any potential loss of statistical data during this time. We’ve put in a fix so that this issue cannot cause an outage in the future.
We identified the issue which was causing sending to be delayed. There was an error with the Open Tracking processing which was causing an issue with our queues. We’ve temporarily disabled Open Tracking while we investigate. All queued messages have now been sent out. We are investigating the Open Tracking issue and will return it to full functionality ASAP.
We’re experiencing a delay in sending. We’re investigating it and will have an update shortly. Any message not immediately sent will be queued for sending shortly.
We just experienced a full outage across all servers in our main Chicago data center. We suspect it is the firewall which failed over to the redundant server. All servers are back online and we are looking into it.
SSL termination issues: After bringing our secondary load balancer back online we noticed issues with SSL termination. A portion of our traffic to the Postmark API may have received “certificate verify failed” during this time. We fixed the issue quickly and all SSL termination is now back to normal. Sorry for the problems.
We are back online. Everything should be back to normal. We are clearing the queue of accepted messages during the offline mode.
Postmark is in offline mode at the moment. All messages will be accepted & delivery may be delayed up to 5 minutes.
All services are running normally.
To recap, at about 3:53pm EST, one of our load balancers died causing a failover for our APIs. Sending email and accessing the API was affected for approximately 30 seconds for a portion of our traffic.
In addition, open tracking and the web UI were affected. Both of these services did not fail over properly and had downtime of about 10 minutes.
We are reviewing our load balancer fail over process and will improve the automation and time to recovery for all of these services. We’re very sorry for any problems this might have caused.
The front-end has been restored to normal working order. We’re double checking all services.
The front-end web ui is experiencing an issue which we’re addressing right now. Email sending is not affected.