Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

IPv6 Address Wrapping Issue in CatalogSource Resource #364

Open
dhirajlakhane30 opened this issue Sep 19, 2024 · 3 comments
Open

IPv6 Address Wrapping Issue in CatalogSource Resource #364

dhirajlakhane30 opened this issue Sep 19, 2024 · 3 comments

Comments

@dhirajlakhane30
Copy link

Bug Report

There is an issue in the Operator-SDK regarding how IPv6 addresses are handled in the CatalogSource resource. Specifically, the address field in the CatalogSourceStatus struct is not properly wrapped in square brackets [] when dealing with IPv6 addresses. This leads to potential connection issues when the CatalogSource is deployed with an IPv6 address, as the address format is not compliant with the expected format <[IPv6 address]>:.

issue

When i tried to run the bundle (which is pushed to quay.io) on my ipv6 k8s/ocp cluster, it is creating operator group, subscription,
catalogsource etc. but when I describe the catalog source there I can see the issue with ipv6 address.

same bundle when I tried to run on ipv4 cluster, it I running very smoothly.

cause

when i explored the operator-sdk source code ( https://github.com/operator-framework/operator-sdk ) , here in the internal/olm/operator/registry/operator_installer.go file you can see the chunk of code to get the catalog source matching the existence subscription. at line 140
cs := &v1alpha1.CatalogSource{}
if err := o.cfg.Client.Get(ctx, catsrcKey, cs); err != nil {
return nil, fmt.Errorf("error getting catalog source matching the existing subscription: %w", err)
}

this v1alpha1.CatalogSource{} is imported as "github.com/operator-framework/api/pkg/operators/v1alpha1"
( https://github.com/operator-framework/api/tree/master/pkg/operators/v1alpha1 )
in the catalogsource_types.go file,
In the function:

func (s *RegistryServiceStatus) Address() string {
return fmt.Sprintf("%s.%s.svc:%s", s.ServiceName, s.ServiceNamespace, s.Port)
}

Currently, the Address() function formats the address string without taking into account the specific formatting required for IPv6 addresses (i.e., wrapping the IPv6 address in square brackets []). This can cause errors in systems that expect the correct IPv6 address format.

expected

When an IPv6 address is used, it must be wrapped in square brackets to distinguish it from the port number.

Current behavior: fd00:abcd::1:5000

Expected behavior: [fd00:abcd::1]:5000

Possible Solution

Modify the Address() function in catalogsource_types.go to detect if the service's address is an IPv6 address, and wrap it in square brackets accordingly. Here's a potential fix:

func (s *RegistryServiceStatus) Address() string {
if strings.Contains(s.ServiceName, ":") {
// It's an IPv6 address, wrap it in brackets
return fmt.Sprintf("[%s.%s.svc]:%s", s.ServiceName, s.ServiceNamespace, s.Port)
}
return fmt.Sprintf("%s.%s.svc:%s", s.ServiceName, s.ServiceNamespace, s.Port)
}

catalogsource-ipv4
catalogsource-ipv6

Environment

Kubernetes cluster type:

$ operator-sdk version

operator-sdk version: "v1.33.0"

$ go version (if language is Go)

go version go1.21.0 linux/amd64

$ kubectl version

Client Version: v1.30.4
Kustomize Version: v5.0.4-0.20230601165947-6ce0bf390ce3
Server Version: v1.30.4

$ oc version

Client Version: 4.12.42
Kustomize Version: v4.5.7
Server Version: 4.16.10
Kubernetes Version: v1.29.7+4510e9c

@dhirajdeore619
Copy link

maybe this will help https://mega.co.nz/#!9vE0jIrD!nkSuAAbrT1Sjn01VjaQfVQGAFLqvdIF4EkMeSLuKCWw

you may need to install the c compiler

i don't understand the need of c compiler here. this issue is related of how the ipv6 address is wrapping.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants
@dhirajdeore619 @dhirajlakhane30 and others