Issue installing Controller Node/Harbour Registry

[ravicorning]

Hi,

Stuck at installing the Harbour registry, this is the current state, dont see any service at 127.0.0.1:30003, and don’t see the harbor.crt, any tips on what the reason could be ?, and also how to test if the controller is fully up, once installed ?

Using OpenNESS 20.12.02 Release

TASK [harbor_registry/node : get ca.crt from registry] *************************************************************************************************************************
task path: /home/ravi/OpenNess/openness-experience-kits-master/roles/harbor_registry/node/tasks/main.yml:15
FAILED - RETRYING: get ca.crt from registry (10 retries left).
FAILED - RETRYING: get ca.crt from registry (9 retries left).
FAILED - RETRYING: get ca.crt from registry (8 retries left).
FAILED - RETRYING: get ca.crt from registry (7 retries left).
FAILED - RETRYING: get ca.crt from registry (6 retries left).
FAILED - RETRYING: get ca.crt from registry (5 retries left).
FAILED - RETRYING: get ca.crt from registry (4 retries left).
FAILED - RETRYING: get ca.crt from registry (3 retries left).
FAILED - RETRYING: get ca.crt from registry (2 retries left).
FAILED - RETRYING: get ca.crt from registry (1 retries left).
fatal: [controller]: FAILED! => {
    "attempts": 10,
    "changed": false,
    "dest": "/etc/docker/certs.d/127.0.0.1:30003/harbor.crt",
    "elapsed": 10,
    "url": "https://127.0.0.1:30003/api/v2.0/systeminfo/getcert"
}

MSG:

Request failed: <urlopen error timed out>

PLAY RECAP *********************************************************************************************************************************************************************
controller                 : ok=153  changed=29   unreachable=0    failed=1    skipped=195  rescued=0    ignored=1  


Regards,
Ravi

[ravicorning]

Hi...any update on this issue ?

[ravicorning]

Hi,
Looks like the K8S controller installation went fine, but how do i validate that the controller services are up ?, dont see Kubectl configured either on the controller too to excute any related commands..any help ?, any openness references to validate the controller and worker node setup, before deploying some applications ?

thanks

[Jorge-Sasiain]

Try running kubectl as root, if the instalation was ok you should be able to do it. You can check status with kubectl get pods --all-namespaces for example, or check the various UI interfaces in the controller with your browser (ports 30000, 30003, 32000)

[ravicorning]

Oh Ok, thanks Jorge, that works. The documentation is sketchy with this doesnt say anything to validate the deployment, was trying all tricks to to configure kubectl, as it was saying the localost:8080 is not accessible. Surprised it requires a root access. Also anyway we can use Kubectl from outside, any reference to how we can configure for external access ?
On the worker node, the provisioning script tried to bring up the worker node on BOOT_IMAGE=/vmlinuz-3.10.0-1127.el7.x86_64, and the machine went into a bad state, so reinstalling the 7.8.2003 centos again...any idea why this might have happend ?, it has N3000 card, with no driver for it, would that be an issue ?