Skip to content

Releases: adobe/koperator

Release 0.13.1-adobe-20210303

03 Mar 13:53
d618fff
Compare
Choose a tag to compare
Pre-release
  • Enable preserveUnknownFields in CRDs (#23)
  • Add Spinnaker annotations fo replace strategy

Release 0.15.1-adobe-public-internal-20210304

04 Mar 13:30
Compare
Choose a tag to compare
  • include public internal required for Kafka migration #22

Release 0.15.1-adobe-20210301

01 Mar 17:02
Compare
Choose a tag to compare

Release 0.13.0-adobe-20201102

02 Nov 15:25
Compare
Choose a tag to compare
  • Improve shutdown script to interract way faster in case of network failure
  • Add uriSans field to kafkaUser generated certs
  • Fix envoyConfig configuration in case of headless service disabled
  • Use GetName() instead of ObjectMeta.Name to avoid NPE
  • Adding log messages when broker pods get deleted
  • Update helm chart to handle cert and webhook generation differently
  • Bump CC version to 2.5.16
  • Fix helm chart lint

Helm chart release chart/kafka-operator/0.4.0-adobe-20201102

02 Nov 15:23
Compare
Choose a tag to compare
Update kafka jmx-exporter metrics rules

- extract broker version
- Extract tls metrics by protocol/cipher
- Extract connection metrics by client software name and version
- Special parser for *Percent beans to also return the MeanRate
and {One,Five,Fifteen}MinuteRate
- parse GAUGES and COUNTERS reported as 3 KV pairs in kafka

Release 0.12.4-adobe-20201027

26 Oct 17:08
Compare
Choose a tag to compare

Envoy multizone (00a9bb4)

  • Dedicated envoy deployment per brokerConfigGroup
  • Add support for dedicated envoy deployment per brokerConfigGroup
    • Use a different label for each envoy deployment
    • The dedicated envoy deployments honors the nodeAffinity and
      the nodeSelector properties se at brokerConfigGroup lelvel.
      This would help placing the Kafka brokers and the ingress
      in the same AZ / rack / DC.
  • Add broker config group to the Status and as pod label
    • this can be used by group-aware resources (such as Envoy)
      during scale up / down, when the Clust Spec is not reflecting
      the real Cluster state
  • Add hostnameOverride per broker config group
    • this will be used to override the ExternalListener DNS
  • Add support to bring your own Load Balancer
    • LB will not be managed by Kafka-Operator.
    • HostnameOverride should be mandatory with BringYourOwnLB
  • ConfigGroupState is optional, for backward compatibility

Update kafka jmx-exporter metrics rules

  • extract broker version
  • Extract tls metrics by protocol/cipher
  • Extract connection metrics by client software name and version
  • Special parser for *Percent beans to also return the MeanRate
    and {One,Five,Fifteen}MinuteRate
  • parse GAUGES and COUNTERS reported as 3 KV pairs in kafka

Helm chart release chart/kafka-operator/0.3.3-adobe-20201027

26 Oct 17:09
Compare
Choose a tag to compare

Envoy multizone (00a9bb4)

  • Dedicated envoy deployment per brokerConfigGroup
  • Add support for dedicated envoy deployment per brokerConfigGroup
    • Use a different label for each envoy deployment
    • The dedicated envoy deployments honors the nodeAffinity and
      the nodeSelector properties se at brokerConfigGroup lelvel.
      This would help placing the Kafka brokers and the ingress
      in the same AZ / rack / DC.
  • Add broker config group to the Status and as pod label
    • this can be used by group-aware resources (such as Envoy)
      during scale up / down, when the Clust Spec is not reflecting
      the real Cluster state
  • Add hostnameOverride per broker config group
    • this will be used to override the ExternalListener DNS
  • Add support to bring your own Load Balancer
    • LB will not be managed by Kafka-Operator.
    • HostnameOverride should be mandatory with BringYourOwnLB
  • ConfigGroupState is optional, for backward compatibility

Release 0.12.4-adobe-20201010

09 Oct 07:50
Compare
Choose a tag to compare

Internal

  • INTERNAL - Sync helm chart CRDs with generated version
  • Add support for dedicated envoy deployment per brokerConfigGroup

Upstream:

  • Upgrade dependencies to latest patch versions
  • Allow to set cruise control log4j configuration via cr
  • Update jmx exporter to version 0.14 and enable cacheing
  • Use the 2.5.13 version of CC and enable kubernetes mode for reporter
  • fix default value in README.md

Helm chart release chart/kafka-operator/0.3.3-adobe-20201010

08 Oct 21:46
Compare
Choose a tag to compare
INTERNAL - Sync helm chart CRDs with generated version

Release 0.12.4-adobe-20200920

18 Sep 22:04
Compare
Choose a tag to compare

Changelog:

Internal

  • INTERNAL - Sync helm chart CRDs with generated version
  • Add support for dedicated envoy deployment per brokerConfigGroup

Upstream:

  • Fix failing test which checks if the alert is handled or not
  • Add github action docker build and use github docker registry
  • Fix alert processed issues which might happened incorrectly if cluster is reconcilling
  • Move incrementby var out of the cycle
  • Fix resizepvc alert command, do not mutate the brokerConfigGroups specs
  • Use the official envoy image instead of the banzai built one
  • Bump CC image to version 2.5.11
  • Bump sarama client version from 2.5 to 2.6
  • Update licence.toml file
  • Update samples to use kafka 2.6.0
  • Add annotation to make istio broker shutdown aware
  • Update dependencies to use the latest patch available