Open Denikin80 opened 7 years ago
I have the same issue, it persist from version 1.7.5(cluster was installed with this version) and now on 1.7.9. kubectl working fine and any other login as well, issue only with GUI access. Issue could be solved with rebooting master and appear again after a day or something. I'm using Google Chrome browser. I would assume that it is due to some auth token expiration or like that.
I used firefox and try on chrome too. Clear browser cache not solve problem.
When i rebooted last 3 master node i can auth Tectonic but there is a message "Oh No! There are something wrong!" or something like that. And there is a screenshot https://ibb.co/j6xOJR https://ibb.co/ciAFXm https://ibb.co/gzQFXm and Events: Liveness probe failed: Get http://10.2.1.12:4180/ping: dial tcp 10.2.1.12:4180: getsockopt: connection refused on worker nodes
Now after some time everything is fine https://ibb.co/e4wdk6 but who gives guarantee that this not happen in the future? Everytime reboot master nodes is not a solution )
two weeks gone - flying normal.
Sorry, so issue just gone without no action? as I still need to reboot master each time, to connect to gui.
Maybe some DNS issue here. I'am using simple DNS round-robin and after shutting down some nodes cannot connect to tectonic console. I'am open another issue on that about best practices on how to make HA for tectonic in case there some nodes are down.
This happens again
I fixed it for my cluster. I did 2 things:
Issue Report Template
Tectonic Version
1.7.9
Environment
What hardware/cloud provider/hypervisor is being used with Tectonic? vmware esxi 5.5
Expected Behavior
login to Tectonic account
Actual Behavior
brouser URL flashed to https://tectonic.maydomain/error?error=invalid_code&error_type=auth and then changed to https://tectonic.mydomain/identity/auth/local?req=aenmaaau4iws6v45kfgykkfua
Reproduction Steps
Other Information
Esterday everything work fine. Today i'am can't login to Tectonic GUI. The password is correct. If enter wrong pass or login - there is red window appears, but in my case i enter valid login and pass? but page reload after that and Tectonic Identity page reappear. Rebooting worker nodes does not solve problem. I still able to communicate via ssh but not GUI. After rebooting worker nodes i see this in console (maybe it is not the issue but maybe helpful): https://ibb.co/bSeHCm
Feature Request
Environment
What hardware/cloud provider/hypervisor is being used with Tectonic?
Desired Feature
Other Information
core@worker1 ~ $ kubectl get nodes -o wide NAME STATUS AGE VERSION EXTERNAL-IP OS-IMAGE KERNEL-VERSION core1.mydomain Ready 4d v1.7.9+coreos.0 Container Linux by CoreOS 1520.8.0 (Ladybug) 4.13.9-coreos
core2.mydomain Ready 4d v1.7.9+coreos.0 Container Linux by CoreOS 1520.8.0 (Ladybug) 4.13.9-coreos
core3.mydomain Ready 4d v1.7.9+coreos.0 Container Linux by CoreOS 1520.8.0 (Ladybug) 4.13.9-coreos
worker1.mydomain Ready 4d v1.7.9+coreos.0 Container Linux by CoreOS 1520.8.0 (Ladybug) 4.13.9-coreos
worker2.mydomain Ready 4d v1.7.9+coreos.0 Container Linux by CoreOS 1520.8.0 (Ladybug) 4.13.9-coreos
worker3.mydomain Ready 4d v1.7.9+coreos.0 Container Linux by CoreOS 1520.8.0 (Ladybug) 4.13.9-coreos