Skip to content

v0.15.0

Latest

Choose a tag to compare

@Despire Despire released this 29 Jun 12:40
300eb16

Claudie v0.15.0

Readme Before Deploying

Autoscaler changes

The autoscaler will no longer be deployed by Claudie. From now it becomes an opt-in paid feature.

Make sure no tasks are scheduled

Before deploying this Claudie release, make sure no tasks are actively being worked on. You can verify this via

kubectl get inputmanifests

All of the input manifests should be in state WATCHING_FOR_CHANGES.

You can also scale down the number of replicas for the manager service to 0 to be sure no tasks are scheduled.

kubectl scale deploy/manager -n claudie --replicas=0

MongoDB

This version of Claudie updates MongoDB.

It is required that before deploying this new version that the number of
replicas of the current deployment for MongoDB is scaled down to 0.

kubectl scale deploy/mongodb -n claudie --replicas=0

After deploying:

  • Check if the version running in the primary replica is 8.3 via
kubectl exec -it <primary-mongo-pod> -n claudie -- mongosh \
  -u <username> -p <password> --authenticationDatabase admin \
  --eval "db.adminCommand({ buildInfo: 1 }).version"
  • Set the feature set to version 8.3
kubectl exec -it <primary-mongo-pod> -n claudie -- mongosh \
  -u <username> -p <password> --authenticationDatabase admin \
  --eval "db.adminCommand( { setFeatureCompatibilityVersion: '8.3', confirm: true } )"

This command must perform writes to an internal system collection. If for any reason the command does not complete successfully, you can safely retry the command as the operation is idempotent.

  • Verify the command successfully updated the feature set.
kubectl exec -it <primary-mongo-pod> -n claudie -- mongosh \
  -u <username> -p <password> --authenticationDatabase admin \
  --eval "db.adminCommand({ getParameter: 1, featureCompatibilityVersion: 1 })"

What's Changed

  • Updated mongo to version 8.3 #2141

  • Cluster-autoscaler has been removed. From now it becomes an opt-in paid feature. #2146

  • Added support spot instances on AWS, Azure, OCI, Verda, GCP #2151, #2149, #2145

  • Faster handling of nodes with unknown status #2150

  • Fixed nil panic on DNS creation failure #2152

  • Updated template parsing for alternative names #2153