first, find the bosh job index of the DEA that has the app instance. hit the stats endpoint for the app [1] using something like:
cf curl /v2/apps/`cf app james-static --guid`/stats
{
"0": {
"state": "RUNNING",
"stats": {
"name": "james-static",
"uris": [
],
"host": "10.10.17.37",
"port": 61035,
"uptime": 442307,
"mem_quota": 1073741824,
"disk_quota": 1073741824,
"fds_quota": 16384,
"usage": {
"time": "2015-03-30 22:02:07 +0000",
"cpu": 4.020839857362368e-05,
"mem": 4595712,
"disk": 8392704
}
}
}
}
this james-static is running on the DEA with the private IP 10.10.17.37
you can compare this to the output of "bosh vms" to find the right DEA bosh job index. "bosh ssh" to that VM.
now you can use the high app CPU instructions to see approximate steps to identify the warden container:
http://docs.cloudfoundry.org/running/troubleshooting/troubleshooting-apps.html
the DEA has an /var/vcap/data/dea_next/db/instances.json
file which should show the app instance guid and associate it with the warden handle.
optional, now you can now use the warden handle with a command like this to get inside the container:
/var/vcap/packages/warden/warden/src/wsh/wsh --socket /var/vcap/data/warden/depot/WARDEN-HANDLE/run/wshd.sock --user vcap