Pivotal Greenplum for Kubernetes 0.7.0 Release Notes
Pivotal Greenplum for Kubernetes is an experimental release of Pivotal Greenplum Database on Kubernetes. Greenplum for Kubernetes 0.7.0 is a minor release that introduces and changes some features; it is based on Pivotal Greenplum 5.15.1.
Refer to the Pivotal Greenplum Database documentation for detailed information on Pivotal Greenplum Database.
Pivotal Greenplum for Kubernetes includes the Pivotal Greenplum Database version identified below:
|Greenplum for Kubernetes Version||Pivotal Greenplum Database Version|
Pivotal Greenplum for Kubernetes 0.7.0 runs on the following platforms:
- Pivotal Container Service (PKS) 1.2 (contains Kubernetes 1.11)
- Google Kubernetes Engine (GKE) Kubernetes 1.11
Additional Kubernetes environments, such as Minikube, can be used for testing or demonstration purposes.
This release introduces the following new and changed features:
- Greenplum for Kubernetes now supports upgrading from a prior release. See Upgrading Greenplum for Kubernetes for instructions.
- You can configure a Greenplum cluster to run on specific Kubernetes workers. This is accomplished using the new
workerSelectorproperty in the Greenplum Operator manifest file.
- Resource group-based resource management is enabled by default in Greenplum clusters. This differs from other Pivotal Greenplum Database installations that use resource queues by default. See Memory and Resource Management with Resource Groups in the Pivotal Greenplum documentation for more information.
- The labels
app=greenplum, greenplumcluster=<name>are now applied to all of the Kubernetes resources that are created for a Greenplum cluster. This includes Kubernetes secrets, PVCs, pods, statefulsets, configmaps, and services. You can use the labels to identify Greenplum cluster resources, for example in commands such as
kubectl get all -l greenplum-cluster=my-greenplum.
- The Greenplum operator no longer records spurious error messages in its log file (obtained using
kubectl logs -l name=greenplum-operator) during initialization.
- Installed tools such as
sshKeyScanshow more detailed command output and fewer warning messages.
To upgrade your existing Greenplum for Kubernetes operator and clusters, follow the instructions in Upgrading Greenplum for Kubernetes.
Note: After upgrading Greenplum for Kubernetes, you can also rollback to the previous version 0.6.0 software if necessary. To rollback to the previous version, use the same upgrade instructions, but delete the version 0.7.0 sofware and cluster, and install and upgrade using the older version 0.6.0 software. Keep in mind that when you upgrade or downgrade any deployed Greenplum cluster, you must first delete the deployed cluster and then re-create it with the desired Greenplum operator software, as discussed in Upgrade a Deployed Greenplum Cluster.
- Greenplum for Kubernetes does not support the built-in SNMP features that are available in Pivotal Greenplum Database.
- Pivotal Greenplum for Kubernetes is an experimental product and is not intended for use in a production environment. Experimental features are subject to change without notice in future releases.
- Greenplum for Kubernetes does not currently support cluster monitoring with Greenplum Command Center. Use system-level monitoring tools such as Prometheus and Grafana until Greenplum Command Center support is available.
- Greenplum for Kubernetes clusters do not support installing Greenplum extensions that use the
gppkgutility). Future releases will include these extensions as part of the distribution, as with MADlib.
- The Greenplum Operator does not yet support changing all attribute values of a deployed Greenplum cluster. In order to change attributes such as the cluster name or segment resource usage, you must delete the cluster and then redeploy using the changed values.
- Certain Greenplum resources that are created during deployment are not yet labeled for convenient management with
kubectl. This includes Greenplum