containers are not being load balanced evenly #284
Description
We noticed that one of our concourse workers is being overloaded with containers, even though there are other workers which have a lot of capacity. Here is the status of the workers
01e85f7c-638f-43e5-9119-854826ba960a 84 linux none
27e89480-4006-4ea5-9771-518a44076113 256 linux none
2e96519b-81fb-446e-bca4-edd7bb29fb78 97 linux none
5e6fd75a-185b-4cb1-8bbb-1299f87df6e1 0 linux vsphere, cork
b4f1522d-3fc3-4ccd-a7c0-d75437aa5418 1 linux aws
c6305523-3110-46ac-996c-4f2788670d97 1 linux aws
f188c382-3540-434a-b651-c36ad1bc2bf0 0 linux vsphere, cork
f4c82e25-ae6c-4435-a6ac-815abe329f68 84 linux none
This results in jobs failing with the error insufficient subnets remaining in the pool.
We are on version concourse/0.68.0 and garden-linux/0.328.0.