As I'm following the development quite a bit, I noticed you are using an official AWS AMI as agent.

I just want to make you aware of an issue I have been debugging with cloudbees stuff for a couple of months until Yoann Dubreuil finally found the issue...

I was having issues to execute long running DB update tasks from within jenkins, everytime the tasks ran into a timeout...

This was caused by a timeout on the NAT gateway.

You can read more about it here:

• http://docs.aws.amazon.com/AmazonVPC/latest/UserGuide/vpc-nat-gateway.html#nat-gateway-troubleshooting-timeout 
• http://docs.aws.amazon.com/redshift/latest/mgmt/connecting-firewall-guidance.html#connecting-firewall-guidance.change-tcpip-settings

The issue could finally be fixed with the following settings (placed in a file in `/etc/sysctl.d`):

net.ipv4.tcp_keepalive_time=200  
net.ipv4.tcp_keepalive_intvl=200  
net.ipv4.tcp_keepalive_probes=5

 
As I'm very interested in using evergreen myself one day, I would love to see this find place in the AWS evergreen setup.

Here is an other article describing the same issue: https://www.vertica.com/blog/configuring-tcp-idle-settings-long-running-idle-sessions/