Andrei Kvapil
Andrei Kvapil
This PR introduces the default instancetypes: - cx1.2xlarge - cx1.4xlarge - cx1.8xlarge - cx1.large - cx1.medium - cx1.xlarge - gn1.2xlarge - gn1.4xlarge - gn1.8xlarge - gn1.xlarge - m1.2xlarge - m1.4xlarge...
Currently someone with access to the cluster just pick a different backing storage class by changing it? This should be fixed by specifying storageClassEnforcement in kubevirt-csi-controller. https://github.com/kubevirt/csi-driver/blob/35836e0c8b68d9916d29a838ea60cdd3fc6199cf/docs/snapshot-driver-config.md?plain=1#L40-L43 This should specified...
## Summary by CodeRabbit - **New Features** - Introduced S3-compatible domain name configuration for SeaweedFS. - Added support for secure HTTPS connections in S3 service. - New `s3` and `cosi`...
Sometimes my ovs-ovn pods can't become ready, with these logs: ``` 2024-10-09T13:50:23.367Z|00051|rconn|WARN|unix:/var/run/openvswitch/br-int.mgmt: connection failed (No such file or directory) 2024-10-09T13:50:23.387Z|00015|rconn(ovn_pinctrl0)|WARN|unix:/var/run/openvswitch/br-int.mgmt: connection failed (No such file or directory) 2024-10-09T13:50:27.367Z|00052|poll_loop|INFO|Dropped 119495 log...
After eviction there are a lot of ContainerStatusUnknown pods especially in kube-ovn namespace. Righ now they can be simple removed ``` kubectl get pod -A | grep ContainerStatusUnknown | awk...
We have to align all alerts to have specific labels https://github.com/alerta/alerta/blob/b7111d23ae9aefb5064ef107976f62e3836217b6/alerta/webhooks/prometheus.py#L49-L57
Helm does not provide CRDs management at all. We need to move all `*/crds ` into separate chart and reconcile it before every operator reconciliaton
We need to add alerts to track the state of HelmReleases
Add more test cases, to test Kuberntes cluster deployment as well any other application from the dashboard https://github.com/aenix-io/cozystack/blob/main/hack/e2e.sh