Open smcavoy-travelbird opened 8 years ago
Sean, the 2.6.6 node source is now asynchronous by default, and ec2 plugin has its own asynchronous behavior, it might be just be delayed in displaying the results. you can also try disabling the asynch nodes feature in the project
When creating a job and selecting a set of nodes to have a step execute on there is an option to refresh the list. I assume this is to query AWS for an updated inventory at that time. Is that not the case?
Are there no options to have more verbose logging?
I found the problem, I did not add an endpoint, I left it blank. Is that expected behaviour?
@smcavoy-travelbird what region are you using? leaving it blank uses the default endpoint, which i think is us-east-1
We have nothing in us-east-1, only eu-west-1/eu-central-1 So this explains it. It would be helpful if that was clearer so I created a PR: https://github.com/rundeck-plugins/rundeck-ec2-nodes-plugin/pull/48
Perhaps the text for the 'endpoint' input field should also have more information there, and/or have the default be shown explicitly so it's clear what it is
I have installed 1.5.2 on 2.6.6 and am not seeing any nodes populating.
I tried a false key and received and error. Fixed it, no nodes. I checked the key is authorised to run describe-instances And I have seen https between rundeck and a AWS IP that was timed with the reload of the node list.
I do not see any errors related to getting the details from AWS. I see no option to increase the verbosity of logging to determine why this is not working.
The only odd thing in the logs I found is below:
Has this behaviour been seen before? Is there anyway to further debug this problem? Thanks,
Sean