We saw another period of worker restarts due to connectivity issues between our OS X infrastructure and our RabbitMQ clusters. Between 16:37 and 17:37 UTC we accrued and processed some additional backlog. At this time things are remaining stable so we're resolving this issue.
Posted Oct 22, 2015 - 17:50 UTC
Monitoring
The workers are back online and there's no backlog. We'll continue to monitor and investigate what caused the workers to go offline. Sorry for the interrruption.
Posted Oct 22, 2015 - 11:51 UTC
Investigating
We're investigating an issue with OS X jobs not running for travis-ci.org and travis-ci.com.