'Jenkins inbound-agent jnlp failing or not spinning up the pod on a kubernetes cluster
I have a Jenkins (3.8.8 installed via helm) instance running on kubernetes (EKS). It needs to pull a repo from git and build the image and push to the container repository.
Most recently when the jnlp (inbound-agent) should spin up to build the repo, it would not even start the pod but instead will throw the following warning / error
Error: Error response from daemon: Minimum memory limit allowed is 6MB
If there will be a chance that it does spin up, I get the following
WARNING: Failed to load winp. Reverting to the default
java.lang.NoClassDefFoundError: hudson/util/ProcessTree$Linux$LinuxProcess
at hudson.util.ProcessTree.get(ProcessTree.java:466)
But then it would still end up getting terminated eventually once some of my steps gets executed (pipeline).
I can't find any resource online that describes the latter (or former really). I've upgraded to a larger EC2 instance to increase the memory but still the same.
Forgot to mention that the namespace also has enabled istio on it. Which has the holdapplicationuntilproxystarts set to true
Sources
This article follows the attribution requirements of Stack Overflow and is licensed under CC BY-SA 3.0.
Source: Stack Overflow
| Solution | Source |
|---|