origin icon indicating copy to clipboard operation
origin copied to clipboard

Published 20 hours ago • verified publisher icon openshift

Build macos `oc` binary with CGO_ENABLED=1 so that it honours /etc/resolver?

Open praveenkumar opened this issue 5 years ago • 18 comments

Version
$ oc version
oc v4.0.0-0.171.0
kubernetes v1.12.4+a532756e37
Steps To Reproduce
  1. Create a cluster locally which doesn't have public dns lookup (you can use crc for it).
  2. Don't append the dns details to /etc/resolve.conf but use /etc/resolver/<whatever_dns> , in our case we use /etc/resolver/testing.
  3. Try to open the web console on the browser which works and it honour the resolver file.
  4. Try to use oc cli to login to cluster, it doesn't work until you add that dns to /etc/resolv.conf
Current Result

User need to add the dns details to /etc/resolv.conf, where most of the app on the Mac does honour the resolver file and works.

Expected Result

Openshift client binary should also honour the resolver directory files.

Additional Information

[try to run $ oc adm diagnostics (or oadm diagnostics) command if possible] [if you are reporting issue related to builds, provide build logs with BUILD_LOGLEVEL=5] [consider attaching output of the $ oc get all -o json -n <namespace> command to the issue] [visit https://docs.openshift.org/latest/welcome/index.html]

praveenkumar avatar Jun 25 '19 05:06 praveenkumar

https://github.com/golang/go/issues/12524

praveenkumar avatar Jul 04 '19 08:07 praveenkumar

especially: https://github.com/golang/go/issues/12524#issuecomment-499525771 Enabling CGO would seem to allow DNS lookups to happen using /etc/resolver.

gbraad avatar Jul 04 '19 08:07 gbraad

After building oc with CGO_ENABLED=1, I can confirm it will use /etc/resolver/ content for DNS resolution

$ cat /etc/resolv.conf
search cdg.redhat.com redhat.com win.redhat.com
nameserver 10.x...
nameserver 10.y...

$ cat /etc/resolver/testing 
port 53
nameserver 192.168.64.8

$ oc config view
apiVersion: v1
clusters:
- cluster:
    certificate-authority-data: DATA+OMITTED
    server: https://api.crc.testing:6443
  name: crc

$ ~/go/bin/oc get nodes
NAME                 STATUS   ROLES           AGE   VERSION
crc-fsjcf-master-0   Ready    master,worker   11d   v1.13.4+9252851b0

The binary from https://mirror.openshift.com/pub/openshift-v4/clients/oc/latest/macosx/oc.tar.gz fails to resolve the DNS name:

$ oc get nodes
Unable to connect to the server: dial tcp: lookup api.crc.testing on 10.38.5.26:53: no such host

cfergeau avatar Jul 05 '19 09:07 cfergeau

https://github.com/kubernetes/kubernetes/issues/23130 upstream issue from kubernetes side.

praveenkumar avatar Jul 05 '19 09:07 praveenkumar

This seems to impact odo as well - does that need to be raised as a separate issue in their tracker?

jankleinert avatar Jul 26 '19 12:07 jankleinert

Note this affects openshift-installer and other binaries

sspeiche avatar Jul 26 '19 12:07 sspeiche

@derekwaynecarr how to handle this? So far we haven't gotten anywhere with this and it affects us on macOS, and we find out the workaround with /etc/hosts we have is a rat's nest, as it is not guaranteed to work due to permissions being modified outside of our control.

gbraad avatar Oct 10 '19 10:10 gbraad

Does kubectl have this same problem too?

sspeiche avatar Oct 10 '19 16:10 sspeiche

I just tried it, and yes it behaves the same, it needs an entry in /etc/hosts and is not making use of /etc/resolver/

cfergeau avatar Oct 11 '19 10:10 cfergeau

@cfergeau Should this be raised upstream? Seems like odo should follow what kubectl does.

Curious what @soltysh thinks

sspeiche avatar Oct 11 '19 12:10 sspeiche

There is no impact to the software delivered, but rather inthe build process as Apple-branded hardware needs to be involved, although this can be solved by using CircleCI.

gbraad avatar Oct 11 '19 13:10 gbraad

Curious what @soltysh thinks

If we want that fixed, this has to be fixed upstream first, imo.

soltysh avatar Oct 14 '19 13:10 soltysh

Issues go stale after 90d of inactivity.

Mark the issue as fresh by commenting /remove-lifecycle stale. Stale issues rot after an additional 30d of inactivity and eventually close. Exclude this issue from closing by commenting /lifecycle frozen.

If this issue is safe to close now please do so with /close.

/lifecycle stale

openshift-bot avatar Jan 12 '20 17:01 openshift-bot

/remove-lifecycle stale

cfergeau avatar Jan 13 '20 11:01 cfergeau

Issues go stale after 90d of inactivity.

Mark the issue as fresh by commenting /remove-lifecycle stale. Stale issues rot after an additional 30d of inactivity and eventually close. Exclude this issue from closing by commenting /lifecycle frozen.

If this issue is safe to close now please do so with /close.

/lifecycle stale

openshift-bot avatar Apr 12 '20 13:04 openshift-bot

/remove-lifecycle stale

cfergeau avatar Apr 14 '20 09:04 cfergeau

Issues go stale after 90d of inactivity.

Mark the issue as fresh by commenting /remove-lifecycle stale. Stale issues rot after an additional 30d of inactivity and eventually close. Exclude this issue from closing by commenting /lifecycle frozen.

If this issue is safe to close now please do so with /close.

/lifecycle stale

openshift-bot avatar Jul 13 '20 12:07 openshift-bot

/lifecycle frozen

cfergeau avatar Aug 04 '20 12:08 cfergeau

This issue is solved in https://github.com/openshift/oc/issues/315

daliborfilus avatar Apr 29 '24 09:04 daliborfilus

/close

praveenkumar avatar May 01 '24 06:05 praveenkumar

@praveenkumar: Closing this issue.

In response to this:

/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.

openshift-ci[bot] avatar May 01 '24 06:05 openshift-ci[bot]