You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is similar to #3267. The log message karpenter generates when it fails to launch a node due to hitting a provisioner memory limit is (masking provisioner names for security):
Could not schedule pod, incompatible with provisioner "*****-provisioner", did not tolerate ******-dedicated=:NoSchedule; incompatible with provisioner "karpenter-provisioner", incompatible requirements, label "********-dedicated" does not have known values; incompatible with provisioner "**********-provisioner", no instance type satisfied resources {"cpu":"1","memory":"50Gi","pods":"1"} and requirements kubernetes.io/arch In [amd64], karpenter.sh/capacity-type In [on-demand spot], node.kubernetes.io/instance-type In [m5.2xlarge m5.4xlarge m5.large m5.xlarge], kubernetes.io/os In [linux], karpenter.sh/provisioner-name In [*******-provisioner], node.umbra.space/exclude-from-external-load-balancers In [true], *******-dedicated In [true] {"commit": "c4a4efd-dirty", "pod": "********/**********"}
I recently ran into a similar issue where the problem was the instance type that was needed was currently not available in the region. The log message was basically the same as the one in this issue - very generic. I had to manually try to launch an instance with that type to see what the real problem was. And like @bwmetcalf pointed out in that #3267, that one has the same problem.
billrayburn
added
v1
Issues requiring resolution by the v1 milestone
chore
Issues that are refactors or code cleanup tasks
and removed
bug
Something isn't working
labels
Mar 29, 2023
Version
Karpenter Version: v0.22.1
Kubernetes Version: v1.24.0
This is similar to #3267. The log message karpenter generates when it fails to launch a node due to hitting a provisioner memory limit is (masking provisioner names for security):
Karpenter used to log using the following line of code which was much clearer in this situation: https://github.com/aws/karpenter-core/blob/0613466c72cfbc836d0df4c0f2bb54c790358978/pkg/apis/v1alpha5/limits.go#L36
Expected Behavior
Karpenter should use a clearer error message when hitting provisioner limits.
Actual Behavior
Karpenter logs an ambiguous error message.
Steps to Reproduce the Problem
Cause karpenter fail to provision a node due to memory limits.
Resource Specs and Logs
See about for logs.
Community Note
The text was updated successfully, but these errors were encountered: