Massive externalNode performance degradation after upgrade to 1.16

[bryder]

We have what looks like a similar problem.

Crippling load after update to foreman 1.17

Problem: Clustered install with 4 foreman nodes behind an F5 load balancer and ~8000 clients running Foreman 1.15 and external posgres database and puppetmasters i.e. foreman nodes only do ENC and UI. Upgraded to 1.17 via 1.16 and everything seemed ok with a small amount of clients but once we started to load up the cluster with client requests again CPU on all nodes went to 100% crippling the cluster. We added multiple new foreman nodes up to 7 in total but it made little difference so event…

Ignore the api build problems in that orinal post.