Having a weird situation where it seems like there’s a 5 minute delay before an Ansible task actually does anything. It takes about 5 minutes 45 seconds before I start seeing any Ansible output from the task. We are not using Ansible git repo caching.
Anyone have any ideas on what’s going on here and how to eliminate that?
Is the repo large in size? The repo caching is to help with folks who have a lot of bandwidth used to clone the git repo.
What’s the provisioned environment? onprem/aws/azure/etc. Shared storage with an HA setup? What version?
It’s a very small repo containing a few Ansible playbooks. The provisioned instance is on-premise VMware (t does the same with any instance, not just one). Morpheus server is a single node, also on-premise VMware. Currently running 5.4.6.
Can you create a test repo, or at least a test playbook with just a ping in it?
- hosts: localhost
tasks:
- ping:
Maybe check the logs right after it completes. Something odd is going on if it’s a small repo.
Agreed. My only thoughts are poor performance to a shared mount for /var/opt/morpheus/morpheus-ui or maybe latency to the repo.
Are other task types pulling from a repo seeing the same delays?
Running this ran very quickly - 3 seconds.
PLAY [localhost] ***************************************************************
skipping: no hosts matched
PLAY RECAP *********************************************************************
Ok, so it’s not ansible itself. Do you have any additional galaxy packages in your task?
We are using a couple - ansible.windows and community.windows (697KB).
Alright, so at this point it seems like it’s specific to the repository. I copied the playbook from one repo to another and it starts very quickly…
I deleted/recreated the repo and deleted/recreated the integration and the 5 minute delay is now gone. 
It sounds like a caching issue. Is your Morpheus in an HA setup with a shared directory?
To address delays in larger ansible galaxy pulls, you could probably install them once into the morpheus-local user, but I haven’t tested that.
In any case, I’m glad that removing it and re-adding cleared the delay.
Very odd. Curious are you connecting via SSH or HTTPS?