search

coreos / fedora-coreos-tracker

Issue tracker for Fedora CoreOS
https://fedoraproject.org/coreos/
264 stars 59 forks source link

Connection Refused while master node for machine-config server url: https://apr-int.ocp4.example.com:22623/config/master #970

Open riteshmishra00 opened 3 years ago

riteshmishra00 commented 3 years ago

Hi Team, We're setting up OKD 4.7 disconnected environment on vmware using baremetal approach Fedora CoreOS used: https://builds.coreos.fedoraproject.org/prod/streams/stable/builds/34.20210821.3.0/x86_64/fedora-coreos-34.20210821.3.0-live.x86_64.iso

Bootstrap node is created or provisioned and machine config service (Port:22623) and Kubernetes API (Port:6443) are started on Bootstrap Node.

Following are the steps followed on creating master nodes

  1. Created a VM with FCOS
  2. Set Static Ip using nmtui
  3. Downloaded master.ign file using coreos-installer command
  4. Validated OCP node and API DNS lookup
  5. Verified the accessibility of machine config service (Port:22623) and Kubernetes API (Port:6443) service with expected SSL certificate and ignition data
  6. Rebooted the master node.

Post reboot we are getting the below error

image

Kindly point us in the right direction to fix this issue

riteshmishra00 commented 3 years ago

Hi Fedora Team,

Can you Please help me on this?

Thanks in advance!!!!!

jlebon commented 3 years ago

We need the full logs to make it easier to debug. Offhand, this looks like networking isn't coming up in the initramfs. After setting up the static IP, are you using --copy-network when running coreos-installer install?

riteshmishra00 commented 3 years ago

After setting up the static IP, are you using --copy-network when running coreos-installer install?

Yes

riteshmishra00 commented 3 years ago

Also passed the coreos entity from kernel but getting the same response

riteshmishra00 commented 3 years ago

coreos.inst=yes coreos.inst.install_dev=sda coreos.inst.image_url= coreos.inst.ignition_url=http://example.com/config.ign ip=10.10.10.2::10.10.10.254:255.255.255.0:core0.example.com:enp1s0:none nameserver=4.4.4.41

dustymabe commented 3 years ago

You may be hitting https://gitlab.freedesktop.org/NetworkManager/NetworkManager/-/issues/729 where ipv6 is winning a race in the initramfs. This should no longer be an issue since we're now running NM via systemd in the initrd, but maybe there's still some corner case issues there.

Can you try to disable IPv6 by adding ipv6.disable to the kernel command line and see if you get past the problem? The issue upstream was fixed (or smoothed over) in NM 1.32, with is in Fedora Linux 35. You could try with the latest next-devel images and see if you have the same problem.

riteshmishra00 commented 3 years ago

@dustymabe Disabled the ipv6.disabled but still getting the same error. Also tried using Fedora Linux 35 but can't see any changes in errors. Every where its same.

dustymabe commented 3 years ago

just checking.. please use ipv6.disable, not ipv6.disabled

If that doesn't work, then we're really going to need the logs from the boot of the machine to debug further. Serial console is typically best so you can copy/paste.