- The lifetime of auto generated TLS certificates is now configurable with the role and roleGroup
config property
requestedSecretLifetime
. This helps reducing frequent Pod restarts (#598).
- BREAKING: Use distinct ServiceAccounts for the Stacklets, so that multiple Stacklets can be deployed in one namespace. Existing Stacklets will use the newly created ServiceAccounts after restart (#594).
- The operator can now run on Kubernetes clusters using a non-default cluster domain.
Use the env var
KUBERNETES_CLUSTER_DOMAIN
or the operator Helm chart propertykubernetesClusterDomain
to set a non-default cluster domain (#574).
- Reduce CRD size from
1.4MB
to96KB
by accepting arbitrary YAML input instead of the underlying schema for the following fields (#548):podOverrides
affinity
- Fix bug where the configuration of the
hbaseRootdir
at the role level is ignored (#584).
- Implement
envOverrides
for HbaseCluster (#550). - Omid test: use 1.1.2, update default port number and raise test timeout (#556).
- An invalid
HBaseCluster
doesn't cause the operator to stop functioning (#[575]).
- test: Remove ZooKeeper 3.8.4 (#558).
- Added support for HBase 2.6.0 with the following changes (#506):
- Added
clusterConfig.authorization
property to support the OPA authorizer. - Configure log4j2 properties.
- Use built-in prometheus metric exporter.
- Added
- Added support for HBase 2.4.18 (#523).
- Bump
stackable-operator
from0.64.0
to0.70.0
(#524). - Bump
product-config
from0.6.0
to0.7.0
(#524). - Bump other dependencies (#527).
- Remove sed calls to replace KRB REALM in configuration. IMPORTANT: This means the operator will ONLY work with HBase versions that are compiled against Hadoop 3.x as of now (#531)
- Support user authentication using Kerberos (#436).
- More CRD documentation (#425).
- Helm: support labels in values.yaml (#441).
operator-rs
0.56.1
->0.57.0
(#425).- Use new label builders (#438).
- Use Snafu instead of thiserror (#438).
- [BREAKING] Removed legacy node selector on roleGroups (#438).
- Remove support for version
2.4.12
(#454).
- Processing of corrupted log events fixed; If errors occur, the error messages are added to the log event (#511).
- Default stackableVersion to operator version (#385).
- Configuration overrides for the JVM security properties, such as DNS caching (#389).
- Support PodDisruptionBudgets (#399).
- Support graceful shutdown (#402).
- Added support for version 2.4.17 (#403).
vector
0.26.0
->0.33.0
(#382, #403).- Use jmx_exporter soft link instead of hardcoded version (#403).
- Fix Zookeeper hbase.rootdir when users point to discovery ConfigMap of ZookeeperCluster rather than ZNode. Print a warning in that case (#394).
- Default
hbase.unsafe.regionserver.hostname.disable.master.reversedns
totrue
, to ensure the names of RegionServers are resolved to their hostnames instead of IP addresses (#418).
- Removed support for 2.4.6, 2.4.8, 2.4.9, 2.4.11 (#403).
- Generate OLM bundle for Release 23.4.0 (#350).
- Missing CRD defaults for
status.conditions
field (#360). - Set explicit resources on all containers (#366, #378).
- Support podOverrides (#371, #373).
- Operator-rs:
0.40.2
->0.44.0
(#349, #366, #375). - Use 0.0.0-dev product images for tests and examples (#351).
- Use testing-tools 0.2.0 (#351).
- Run as root group (#359).
- Added kuttl test suites (#369)
- Fix missing quoting of env variables. This caused problems when env vars (e.g. from envOverrides) contained a whitespace (#356).
- Fix
hbase.zookeeper.quorum
to not contain the znode path, instead pass it viazookeeper.znode.parent
(#357). - Add
hbase.zookeeper.property.clientPort
setting, because hbase sometimes tried to access zookeeper with the (wrong) default port (#357). - Fix test assert by adding variable quoting (#359).
- Increase the size limit of the log volume (#375).
- Deploy default and support custom affinities (#322).
- OLM bundle files (#333).
- Extend cluster resources for status and cluster operation (paused, stopped) (#336).
- Cluster status conditions (#337).
- [BREAKING]: Consolidated top level configuration to
clusterConfig
(#334). - [BREAKING] Support specifying Service type.
This enables us to later switch non-breaking to using
ListenerClasses
for the exposure of Services. This change is breaking, because - for security reasons - we default to thecluster-internal
ListenerClass
. If you need your cluster to be accessible from outside of Kubernetes you need to setclusterConfig.listenerClass
toexternal-unstable
(#338). operator-rs
0.36.0
->0.40.2
(#334, #336, #339, #340).- Use
build_rbac_resources
from operator-rs. This renames thehbase-sa
ServiceAccount tohbase-serviceaccount
(#340).
- Avoid empty log events dated to 1970-01-01 and improve the precision of the log event timestamps (#339).
- [BREAKING]: Removed top level role/role group config (#334).
- Log aggregation added (#294).
- [BREAKING] Use Product image selection instead of version.
spec.version
has been replaced byspec.image
(#282). - Updated stackable image versions (#275).
operator-rs
0.24.0
->0.30.2
(#277, #293, #294).- Set runAsGroup to 1000 rather than 0 (#283).
- Fixed:
selector
in role groups now works. It was not working before (#293)
- Startup probe created and thresholds in liveness and readiness probes fine-tuned (#193).
- Include chart name when installing with a custom release name (#209, #210).
- Orphaned resources are deleted (#215).
- Fix HBase-shell start failure (#218).
- Add integration tests and usage documentation for Phoenix (#221).
- Added OpenShift compatibility (#232)
- Support for HBase 2.4.9 (#133).
- Support for HBase 2.4.11 (#148).
- Support for HBase 2.4.12 (#197).
- Use cli argument
watch-namespace
/ env varWATCH_NAMESPACE
to specify a single namespace to watch (#137). - Writing a discovery config map containing
hbase-site.xml
with thehbase.zookeeper.quorum
property (#163).
operator-rs
0.12.0
->0.15.0
(#137, #153).- Now using HDFS discovery config map instead of hdfs name node config map (#153)
- BREAKING: Consolidated CRD - discovery config maps now top level, removed several
HbaseConfig
options (can still be overridden) (#162):hbaseManagesZk
: defaults to falsehbaseClusterDistributed
: defaults to true
- [BREAKING] Specifying the product version has been changed to adhere to ADR018 instead of just specifying the product version you will now have to add the Stackable image version as well, so
version: 3.5.8
becomes (for example)version: 3.5.8-stackable0.1.0
(#179)
- Reconciliation errors are now reported as Kubernetes events (#127).
operator-rs
0.10.0
->0.12.0
(#127).- Migrated to StatefulSet rather than direct Pod management (#110).
operator-rs
:0.3.0
(#18)