Node IP and Route v4 Interface Dropdowns Empty in Kubernetes Advanced Settings

Description

In the Kubernetes Settings both the Node IP and route v4 interface dropdowns are empty (well they have a - ).

I have been messing around with k3s a lot since and I wanted all the apps to run on one subnet/interface and everything else to run on another. I had this working, but ended up having to do it outside the GUI since I had been screwing around and "lost" those options. It got to the point where I had stupidly deleted the interface config that k3s was running on, which made it listen on all interfaces, which somehow made it kill the web UI, but all the containers were running on both subnets. I thought I could fix it by modifying /etc/rancher/k3s/config.yaml since it had the nodeIP as 0.0.0.0 changing that to 192.168.3.7 rendered both that and 192.168.1.7 inaccessible for some reason.

I decided to do a reinstall, but kept the ix-applications dataset, upon starting up k3s again with a fresh install evenything went to hell again. So I did another fresh install and before I started k3s I deleted the ix-applications dataset so everything would be fresh.

The dropdowns were still empty on a clean install. I had installed the latest public beta during one of the aforementioned reinstalls and noticed that those fields were populated with the correct values, so it seems that something between the latest beta and the latest nightly caused it to stop working.

version: TrueNAS-SCALE-21.09-MASTER-20210912-212921

Problem/Justification

None

Impact

None

Activity

Show:

Brandon Golway

October 5, 2021 at 8:41 PM

Hey it appears that this was fixed somewhere between 21.09-MASTER-20210930-112933 and 21.10-MASTER-20211004-172926, because it was there before I updated. Awesome!

Waqar

October 5, 2021 at 5:33 PM

@Brandon Golway can you please update to latest nightlies and confirm if the issue exists ? Also please don't modify anything directly on the filesystem. If the issue still exists on the latest nightlies, please upload a newer debug and share the output of the following command:

I have not been able to reproduce this locally. Thanks!

Brandon Golway

September 14, 2021 at 7:37 PM

I just noticed that starting in the recent beta and configuring the cluster will work, but once you switch to the nightly train the networking largely breaks. I had about 10 deployments in Beta, but then switched to the latest nightly and all but 2 were stuck on deploying citing either issues creating a volume or the network not being ready.

While on the nightly I decided to nuke all my deployments and the ix-applications dataset itself and start over. When I did, the advanced settings section no longer contains the pool dropdown and the other dropdowns are empty, but deployments now work as expected and configured with the intended interface/ip

Resize issue view side panel

Cannot Reproduce

Details

Assignee

Triage Team

Reporter

Brandon Golway

Labels

Impact

High

Time remaining

Fix versions

N/A

Affects versions

Master - TrueNAS Nightlies

Priority

Low

Katalon Platform

Created September 13, 2021 at 4:38 AM

Updated July 6, 2022 at 8:56 PM

Resolved October 6, 2021 at 7:46 AM