bootstrap: More details for flux-system context deadline exceeded
#4411
Comments
|
I believe it could be a very difficult to reproduce issue. |
|
Flux waits 5 minutes by default for the bootstrapping to succeed which is enough time in most cases. If you need more time, use the |
|
Oh... I am stupid one... it was cluster misconfiguration. I used an improper configuration for CoreDNS. Was: apiVersion: v1
kind: ConfigMap
metadata:
name: coredns
namespace: kube-system
data:
Corefile: |
.:53 {
errors
health
ready
kubernetes cluster.local in-addr.arpa ip6.arpa {
pods insecure
fallthrough in-addr.arpa ip6.arpa
}
prometheus :9153
cache 30
loop
reload
loadbalance
}The proper one: apiVersion: v1
kind: ConfigMap
metadata:
name: coredns
namespace: kube-system
data:
Corefile: |
.:53 {
errors
health
ready
kubernetes cluster.local in-addr.arpa ip6.arpa {
pods insecure
fallthrough in-addr.arpa ip6.arpa
}
prometheus :9153
cache 30
loop
reload
loadbalance
forward . 8.8.8.8 <----- add this line
}Now everything is working. It is very interesting that the error message from flux bootstrap does not give any idea about DNS issues. Finally everything is working: |
|
The second finding is that |
|
|
|
@stefanprodan Hi! I completely agree, but it would be nice to get some additional info from FluxCD side to debug the issue :-) I understand that it was my fault to forget some config settings, but if we think logically - if there is no access to main repo defining |
|
Flux can be configured to sync from hundreds of Git repos, Helm repos, OCI registries, S3 buckers, etc Why would we consider that Flux is not working just because some sync fails, it can be transient network error or some external service having an outage. Flux CLI gives you all the tools you need to diagnose such issues |
gecube
changed the title
|
Flux can't diagnose your Kubernetes cluster for misconfiguration but assumes a functioning cluster. There's a lot of guides and tooling out there to diagnose cluster misconfigurations, e.g. the official Kubernetes documentation. |
|
@makkes I completely agree and understand that there is no need to check for transient issues (I thought about them). And I share the opinion that FluxCD is not a diagnose tool for the cluster. The issue is that as flux user I want to get the conscious error messages. Get any idea that you have a wrong config for cluster from If there would be error message like "dns resolving error", or "failed to checkout and determine revision", or something like the same verbosity level, I'd be very happy. Not just simple "context deadline exceeded" Thanks for the attention. |
stefanprodan
changed the title
flux-system context deadline exceeded
stefanprodan
added
enhancement
Describe the bug
Good day colleagues!
Please check the log below:
Looks like that I was hit by rate limit.
Steps to reproduce
Prepare the empty cluster on single node with kubeadm + cilium.
Before installation of FluxCD:
After installation of FluxCD:
Expected behavior
Add more timeouts and checks to finish installation from the first try and return the proper successful exit code.
Screenshots and recordings
No response
OS / Distro
Ubuntu 22.04
Flux version
v2.1.2
Flux check
Git provider
gitlab saas
Container Registry provider
No response
Additional context
No response
Code of Conduct
The text was updated successfully, but these errors were encountered: