4 ms·

yep, we were using Amazon ECS-Optimized Amazon Linux AMI, docker and agent should've matched there. It was strange behavior, either all agents were connected o

by velkyk 10y ago

yep, we were using Amazon ECS-Optimized Amazon Linux AMI, docker and agent should've matched there. It was strange behavior, either all agents were connected or most of them were disconnected at the same time...

maslam 10y ago
Yes. This is pretty much what we saw.
robbles 10y ago
How did you first discover the bug? Did your deploys just start failing in an obvious way, or did your instances disappear (from ECS's perspective)? I'm wondering if Elastic Beanstalk is hiding this from us in some way. Or maybe just has a workaround built in?