Elastic Cloud on Kubernetes 1.8.0
Breaking changes
The operator does not rely on the $PRE_STOP_MAX_WAIT_SECONDS
environment variable to control <<{p}-prestop,Elasticsearch Pods PreStop hook>> anymore. Use $PRE_STOP_ADDITIONAL_WAIT_SECONDS
instead (defaults to 50 seconds).
- Note ECK will drop support for Openshift 3.11 starting ECK version 2.0.
Enhancements
- Drop "properties: {}" from empty-v1alpha1-patch {pull}4813[#4813]
- Remove beta label from Helm documentation {pull}4802[#4802] (issue: {issue}4796[#4796])
Bug fixes
- Fix Agent volumes when an association has no CA {pull}4833[#4833] (issue: {issue}4832[#4832]
- Fix handling CAs in Agent controller {pull}4807[#4807] (issue: {issue}4790[#4790])
- Improve Elasticsearch cluster health API timeout handling {pull}4803[#4803] (issue: {issue}4507[#4507])
- Agent: fix Elasticsearch association error message {pull}4792[#4792]
Misc
- Update module gopkg.in/yaml.v1 to v2 {pull}4788[#4788]
- Update module github.com/fsnotify/fsnotify to v1.5.0 {pull}4778[#4778]
- Update golang Docker tag to v1.17.0 {pull}4766[#4766]
- Update module sigs.k8s.io/controller-runtime to v0.9.6 {pull}4747[#4747]
- Update module go.uber.org/zap to v1.19.0 {pull}4742[#4742]
- Update module sigs.k8s.io/controller-tools to v0.6.2 {pull}4697[#4697]