Setup/configure constantly rerunning after upgrading to Ubuntu 18.04/OpsWorks agent 4040. Also unable to upload chef.logs to s3. about opsworks_ruby

[jhirn]

Upon further investigation, this is due to the utility (sidekiq) instances running OOM causing a setup/configure. Not sure exactly what logs to check for when an instance reboots itself in ops works but it's clear from the moniitoring tab somethings is running away. Going to try the sidekiq recommendations and jemalloc flavors of 3.1.2. Wish me luck yall.

It seems that configure fans out to other instances and is an afer_depoy hook which is different then the ancient version of opsworks_ruby we were using originally. I just need to solve these sidekiq memory issues and I should be golden but appreciate the sounding board and all of the great work on opsworks_ruby @ajgon!.

[l33z3r]

I am having a similar issue. The sidekiq instances are running OOM because monit keeps starting new instances of sidekiq. It somehow cannot detect that it has successfully started a sidekiq instance and then it continues to start more instances. If you run monit status you will see that it is failing, however the processes are actually being created.