Release v0.2

Changelog for v0.2.0 - 2020-02-13

Notable changes

#34 & #35 Clusters can be scaled up and down

Etcd clusters can be scaled up or down by modifying the .spec.replicas field in the EtcdCluster resource. You can also use kubectl scale to do this.
#145 Perform version upgrade by rolling update of peers

The EtcdCluster resource now has a .spec.version field for specifying the version of etcd to use. If the field is updated to a higher value, the operator will perform a rolling version upgrade.
#136 EtcdCluster supports specifying affinity in pod template

The EtcdCluster resource now has a .spec.podTemplate.affinity field for specifying affinity rules on the underlying etcd pods. This allows users to enforce rules such as scheduling each etcd pod on a different host to enable high-availability clusters.
#125 Set resource requests and limits on Etcd peer containers

The EtcdCluster resource now has a .spec.podTemplate.resources field for specifying resource requests and limits on the underlying etcd pods. This allows users to control how much resources the etcd pods can use.
#110 Support passing annotations through to the underlying pods

The EtcdCluster resource now has a .spec.podTemplate.metadata.annotations field for specifying annotations to be passed through to the underlying etcd pods. These annotations can be used to setup scraping of metrics from the etcd pods with Prometheus.

Everything else

#156 Use pflag
#153 Add a --version flag and log the version before starting the operator
#152 Use ETCDCTL_API=3 in e2e tests
#150 Use controller-tools 0.2.5
#151 Update all dependencies to latest patch versions
#148 Pin etcd client version
#142 Redeploy the etcd-cluster-operator before each E2E test run
#120 Allow scaled-down clusters to be scaled back up (v2)
#121 Patch rather than Update EtcdCluster.Status to avoid conflict errors
#123 Fix some race conditions
#135 Download kubebuilder 2.2.0 assets for tests
#96 Backup docs
#133 Use latest 3.2 patch version
#129 FAQ - explain why we have not added Readiness Probes or Liveness Probes.
#128 Retry the number-of-peers test
#131 Fix broken release image build due to shell invocation
#118 Backup schedule race fix
#122 Update to controller-runtime 0.4.0
#124 Update to controller-tools 0.2.4
#111 Remove insecure kube_rbac_proxy
#95 Implement EtcdBackupSchedule
#103 Rewrite operations guide
#65 Implement backup resource controller
#94 Add debug container docker build option
#88 EventuallyInCluster: a helper to run a command in the cluster and wait for it
#91 Shorter tick time in tests
#89 Ensure that the stopped channel gets closed
#78 Make it easier to re-use an existing Kind cluster for E2E tests
#81 Documentation for deleting a cluster and restoring a deleted cluster.
#80 Set a 1 second timeout for the Etcdclient
#73 Run Circle jobs in parallel
#84 Reduce the volume size in test clusters to 1Mi
#85 Save the kind logs after the e2e tests
#86 Fixups for cluster scale-up