Release 2.0.1 - Bug Fix

this release fixes a bug in the pgo client that prevented a namespace other than default from being used.

Release 2.0

New features include:

• no more TPRs!, the PostgreSQL Operator is now based on CRDs (Custom Resource Definitions), this means the 1.7.0 version forward will not work on older Kubernetes versions that do not support CRDs (e.g. Kube 1.6.X) but does support Kube 1.7 and forward
• clone command is dropped, a future implementation is planned that performs the same feature
• starting with version 2.0, the pgo configuration file uses camel case instead of ALL CAPS
• works with the latest Crunchy Containers 1.6.0 release (PostgreSQL 10.0 and 9.6)
• prior TPR based version is in the 1.5.X branch but will be deprecated in the future

Release 1.5.2

a bug fix release:

• fixed a bug dealing with tainting of nodes, the bug forced you to have all your Kube nodes schedulable

Release 1.5.1

Significant changes in this release include:

• pgo label command - an initial implementation that lets you add user defined labels to clusters
• pgo user command - an initial implementation that lets you add users to clusters and manage passwords
• support for selectors in various commands allowing you to perform changes across a set of clusters
• pgo create cluster adds a --series command flag that lets you create a series of clusters off a base name
• initial support for password aging

Release v1.5

This release includes a few new features and bug fixes over the 1.4 release:

• ability to specify a storageclass to be used when creating the master and replica, this allows for dynamic storage provisioning to be utilized in Postgres cluster deployment
• support for node affinity, this allows users to specify a specific Kube node they want the Postgres master to be deployed to, if not specified, normal Kube scheduling will occur
• support for deploying a Postgres replica to a different node than where the Postgres master is deployed to, this feature allows for better Postgres HA support
• converted over to using godep for the golang package management used for building the operator binaries
• upgraded the code to use the k8s client-go version 3.0.0 API, this allows for Kube 1.6/1.7 feature support

Release 1.4

Major changes in this release include:

• bug fixes
• storage spec definitions in the configuration to better support different forms of PVC allocation
• policy feature - users can define and apply SQL-based policies against a cluster
• refactor of replica deployments, now each replica is its own Kube Deployment, this supports more advanced configurations in the future

Release 1.3.2

a bug fix release including:

• a fix for replica PVC specification that enables shared and non-shared volume types to be specified via configuration files, updated documentation to reflect this change

Release 1.3.0

Updates in this release include:

• added a 'pgo test' command which runs a simple SQL query test against a cluster's services
• added a bash completion file for the pgo command, bash users will appreciate this
• fixed a bug in cluster restore, existing secrets are now copied and used by the restored database, the secret-from command line flag is introduced to support this
• added a 'pgo clone' command which will create a cloned copy of an existing cluster
• altered run.sh script to support running on Openshift Origin 1.5.1, documented Openshift usage and reqts

Release 1.2.0

This is a major release and includes the following updates:

• operator-conf ConfigMap is used to store operator templates
• bug fixes
• capture Postgres Full Version for databases and clusters
• support --version command option for pgo show database and pgo show cluster to filter results based on Postgres Full Version
• removed database concept in favor of just a cluster which can have 0 or more replicas, much cleaner and supports more features in the future
• replaced clear text passwords with secrets which get generated for each cluster
• changed CLI command, to update a cluster, you run 'pgo update mycluster'
• changed CLI command, to backup a cluster, you run 'pgo backup mycluster'
• for command 'pgo show pvc', added a flag called '--pvc-root' which lets you specify where in the PVC you want the listing to be based on, if not specified, the value defaults to '/'
• added better CLI error messages
• fixed a bug in 'pgo show pvc' where logs were not being displayed
• added a status field to the backup TPR, that field is marked complete when the backup job is finished
• add '--ccp-image-tag' command flag for 'pgo create cluster' and 'pgo update' commands, this allows users to override the config file setting
• add scaling command, 'pgo scale mycluster --replica-count=2' will scale up or down the replicas

Release 1.1.0

This release includes the following changes from the 1.0.0 release:

• bug fixes
• updated documentation and examples
• tested on Kube 1.6
• upgrade feature - you can now perform minor and major postgres upgrades
• fixed to run in a namespace of your choosing