IP ranges no longer work

[viceice]

MetalLB Version

0.13.12

Deployment method

Charts

Main CNI

calico

Kubernetes Version

1.27.8

Cluster Distribution

k3s

Describe the bug

IP ranges no longer work on IP pools

To Reproduce

apiVersion: metallb.io/v1beta1
kind: IPAddressPool
metadata:
  name: local
  namespace: metallb-system
spec:
  addresses:
  - 172.30.2.70-172.30.2.99
  autoAssign: false

Apply this pool an try to use 172.30.2.70 from that pool and get the not allowed error on the service.

Expected Behavior

Ranges should work like they did some versions before.

Additional Context

Workaround: use cidr ips

apiVersion: metallb.io/v1beta1
kind: IPAddressPool
metadata:
  name: local
  namespace: metallb-system
spec:
  addresses:
  - 172.30.2.70/31
  - 172.30.2.72/29
  - 172.30.2.80/28
  autoAssign: false

I've read and agree with the following

• I've checked all open and closed issues and my request is not there.
• I've checked all open and closed pull requests and my request is not there.

I've read and agree with the following

• I've checked all open and closed issues and my issue is not there.
• This bug is reproducible when deploying MetalLB from the main branch
• I have read the troubleshooting guide and I am still not able to make it work
• I checked the logs and MetalLB is not discarding the configuration as not valid
• I enabled the debug logs, collected the information required from the cluster using the collect script and will attach them to the issue
• I will provide the definition of my service and the related endpoint slices and attach them to this issue

[viceice]

I'll provide logs later if required, but can't currently access them since i wont destry our production system again.

[superherointj]

I confirm issue. IP range wasn't working for me either.

CIDR workaround worked. Thanks for reporting.

My set-up:
MetalLB from helm-chart: metallb@0.13.12.
Host NixOS, K3s v1.28.4+k3s2, CNI is default for K3s (I believe, flannel, vxlan).

[shimritproj]

I would like to work on it.

[oribon]

thanks! assigned

[shimritproj]

You just need to modify the "autoAssign" field, and this will work for you.

apiVersion: metallb.io/v1beta1
kind: IPAddressPool
metadata:
  name: local
  namespace: metallb-system
spec:
  addresses:
  - 172.30.2.70-172.30.2.99
  autoAssign: true

@viceice

[viceice]

You just need to modify the "autoAssign" field, and this will work for you.

apiVersion: metallb.io/v1beta1
kind: IPAddressPool
metadata:
  name: local
  namespace: metallb-system
spec:
  addresses:
  - 172.30.2.70-172.30.2.99
  autoAssign: true

@viceice

i don't want to enable auto assign. I've now converted all ip ranges to use cidr format.

[shimritproj]

i don't want to enable auto assign. I've now converted all ip ranges to use cidr format.

Okay, I will check it and update here. @viceice

[shimritproj]

Hi, I was able to reproduce the bug on v0.13.12 but not on main, which makes me think this was probably fixed in some commit in between (I'm not totally sure which one). I'll modify the existing tests to include the autoAssign:false case so we won't regress again as I can see it is missing. thanks :) @viceice

[viceice]

thanks, when can we expect a new release?

[fedepaol]

thanks, when can we expect a new release?

sometime next week (I started preparing the rel notes #2261) . Please make sure you don't see the problem deploying from main if you cna

[fedepaol]

Closing this issue based on the above discussion. @viceice feel free to reopen if 0.14.3 (or main) do not work.