Week 25 routine - Githubissues

kiwix / operations

Kiwix Kubernetes Cluster

http://charts.k8s.kiwix.org/

5 stars 0 forks source link

Week 25 routine #94

Closed kiwixbot closed 1 year ago

kiwixbot commented 1 year ago

Infra

[x] Check nodes free space
```
df -h / && df -h /data
```
[x] Nodes and worker system upgrades
```
apt update && apt upgrade
```
[x] Ensure all borg repositories are being updated
[x] Check Pod errors
```
k get pods -A -o wide|grep Error
```

[x] Check Pod restarts

k get pods -A -o wide | pyp -i 'print("\n".join([line for line in l if re.split(r"\s+", line)[4] != "0"]))'

[x] Check if k8s should/could be upgraded

curl -s -H "X-Auth-Token: $SCW_SECRET_KEY" https://api.scaleway.com/k8s/v1/regions/fr-par/clusters/$KIWIX_PROD_CLUSTER|jq ".version,.upgrade_available"
curl -s -H "X-Auth-Token: $SCW_SECRET_KEY" https://api.scaleway.com/k8s/v1/regions/fr-par/versions|jq ".versions[].name"

[x] Upgrade k8s if applicable

Stats

[x] Ensure download.kiwix.org stats are being recorded
[x] Check whether matomo should be upgraded

Projects

[x] UptimeRobot has no alert
[x] youzim.it backlog is reasonable
[x] No systematic failure in tasks.
[x] PRs awaiting your review

Security

[x] Analyze/merge dependabot PRs

Note: this is an automatic reminder intended for the assignee(s).

rgaudin commented 1 year ago

This is an ongoing issue with the system node (bastion). Since last week's reboots the network layer (k8s) is not working properly. Several kube-system pods are not working: node-problem-detector, kilo, nginx-ingress, kube-proxy

It doesn't have much consequence as no service currently runs on this node but should it be any other node it would be a terrible mess.

As what seems to be the problem is now invalidated IP/names, I might bring myself to delete the node and re-register to the cluster but I'd prefer a smarter solution.