mybinder.org is down with 503 Service Temporarily Unavailable

[rgaiacs]

The deployment to the server running https://mybinder.org is failing, see https://github.com/jupyterhub/mybinder.org-deploy/actions/runs/17237231094/job/48906168149.

Until this is fixed, you can access from https://gesis.mybinder.org/.

[rgaiacs]

kubectl --namespace=prod get pod

returns

NAME                                                READY   STATUS             RESTARTS          AGE
analytics-publisher-8668fc8996-c7zlx                0/1     CrashLoopBackOff   619 (2m41s ago)   3d2h
build-cache-prune-29270235-dxhbc                    0/1     Error              4 (55s ago)       103s
federation-redirect-678bb89869-wm6sn                0/1     CrashLoopBackOff   19 (4m53s ago)    79m
federation-redirect-678bb89869-xjmbr                1/1     Running            20 (5m4s ago)     78m
gcs-proxy-868b66cbd8-9l82v                          1/1     Running            4 (3d2h ago)      3d2h
matomo-55b8f4b44b-5g54j                             3/3     Running            0                 3d2h
matomo-55b8f4b44b-7ltsn                             3/3     Running            0                 3d2h
matomo-mysqld-exporter-556c84949b-d5bkz             2/2     Running            0                 3d2h
minesweeper-lzx47                                   1/1     Running            0                 3d2h
prod-cryptnono-74w8c                                2/2     Running            0                 3d2h
prod-grafana-7b5b9f85ff-tf4m7                       1/1     Running            1 (19m ago)       5h20m
prod-ingress-nginx-controller-f8f8bbc6d-b47gz       1/1     Running            0                 3d2h
prod-ingress-nginx-controller-f8f8bbc6d-d6td6       1/1     Running            0                 3d2h
prod-ingress-nginx-controller-f8f8bbc6d-sdh4j       1/1     Running            0                 3d2h
prod-ingress-nginx-defaultbackend-7b7c684c4-xjddm   1/1     Running            0                 3d2h
prod-kube-state-metrics-69dcbc8df4-7cj52            1/1     Running            1 (3d2h ago)      3d2h
prod-prometheus-node-exporter-fxclv                 1/1     Running            0                 3d2h
prod-prometheus-server-7b8d875cbd-wqgwv             2/2     Running            0                 3d2h
proxy-patches-9886b6667-xq9fw                       1/1     Running            0                 3d2h
redirector-74c8c584c8-jzf5l                         1/1     Running            0                 3d2h
static-688ffb87f9-56bvd                             1/1     Running            0                 3d2h

kubectl --namespace=prod logs federation-redirect-678bb89869-xjmbr

Removing host curvenote with 0 weight
Removing host gke with 0 weight
Removing host hetzner-2i2c with 0 weight
Removing host ovh2 with 0 weight
No prime host configured!

[rgaiacs]

There is any why to configure Helm to check if federationRedirect.hosts.*.prime is true? This would catch the error earlier and avoid mybinder.org go into 503.

[manics]

Another option is to get rid of the prime config, and use whichever host has the highest weight?

[rgaiacs]

https://mybinder.org/ is back.

use whichever host has the highest weight?

Looks good to me. I will keep this issue open as a reminder to implement this change.