Skip to content

Latest commit

 

History

History
186 lines (148 loc) · 4.4 KB

MIGRATION.md

File metadata and controls

186 lines (148 loc) · 4.4 KB
Raw

Migration

Migrating from zeebe-full-helm

If you're running an earlier zeebe-full-helm chart release, and you want to migrate to camunda-platform you need to adjust your values.yaml file, since the dependencies (sub-charts) have been renamed.

Old Chart Names New Sub-Chart Names
zeebe-cluster-helm zeebe
zeebe-operate-helm operate
zeebe-tasklist-helm tasklist

Some properties have been renamed or removed, but also better documented. Please check the README for more details.

Example:

Old:

global:
  zeebe: "{{ .Release.Name }}-zeebe"

tasklist:
  enabled: false

zeebe-cluster-helm: {}

zeebe-operate-helm: {}

zeebe-tasklist-helm: {}

New:

global:
  zeebeClusterName: "{{ .Release.Name }}-zeebe"

zeebe: {}

operate: {}

tasklist: {}

Migrating from zeebe-cluster-helm

Be aware there is no longer a zeebe only helm chart. If you want to migrate to camunda-platform you have to move most of the properties under the zeebe object.

Note: There is also a new sub-chart for the standalone gateway. This means some configurations need to be done now on the gateway chart.

For more details please check the README, which documents all possible values for each chart.

Example:

Old:

image:
  repository: camunda/zeebe
  tag: SNAPSHOT
  pullPolicy: Always

clusterSize: 3
partitionCount: 3
replicationFactor: 3
cpuThreadCount: 4
ioThreadCount: 4

gateway:
  replicas: 1
  logLevel: debug
  env:
    - name: ZEEBE_LOG_APPENDER
      value: Stackdriver
    - name: ZEEBE_LOG_STACKDRIVER_SERVICENAME
      value: zeebe-gateway
    - name: ZEEBE_LOG_STACKDRIVER_SERVICEVERSION
      valueFrom:
        fieldRef:
          fieldPath: metadata.namespace
    - name: ATOMIX_LOG_LEVEL
      value: INFO
    - name: ZEEBE_GATEWAY_MONITORING_ENABLED
      value: "true"
    - name: ZEEBE_GATEWAY_THREADS_MANAGEMENTTHREADS
      value: "4"

JavaOpts: >-
  -XX:MaxRAMPercentage=25.0
  -XX:+ExitOnOutOfMemoryError
  -XX:+HeapDumpOnOutOfMemoryError
  -XX:HeapDumpPath=/usr/local/zeebe/data
  -XX:ErrorFile=/usr/local/zeebe/data/zeebe_error%p.log
  -Xlog:gc*:file=/usr/local/zeebe/data/gc.log:time:filecount=7,filesize=8M

# Environment variables
env:
  - name: ZEEBE_BROKER_EXECUTION_METRICS_EXPORTER_ENABLED
    value: "true"
  - name: ATOMIX_LOG_LEVEL
    value: INFO

# RESOURCES
resources:
  limits:
    cpu: 5
    memory: 12Gi
  requests:
    cpu: 5
    memory: 12Gi

New:

Be aware you need to disable tasklist and operate, since they are enabled per default.

As described earlier the gateway object was moved into an own chart, called now zeebe-gateway.

global:
  image:
    repository: camunda/zeebe
    tag: SNAPSHOT
    pullPolicy: Always

zeebe:
  clusterSize: 3
  partitionCount: 3
  replicationFactor: 3
  cpuThreadCount: 4
  ioThreadCount: 4

  JavaOpts: >-
    -XX:MaxRAMPercentage=25.0
    -XX:+ExitOnOutOfMemoryError
    -XX:+HeapDumpOnOutOfMemoryError
    -XX:HeapDumpPath=/usr/local/zeebe/data
    -XX:ErrorFile=/usr/local/zeebe/data/zeebe_error%p.log
    -Xlog:gc*:file=/usr/local/zeebe/data/gc.log:time:filecount=7,filesize=8M

  env:
    - name: ZEEBE_BROKER_EXECUTION_METRICS_EXPORTER_ENABLED
      value: "true"
    - name: ATOMIX_LOG_LEVEL
      value: INFO

  resources:
    limits:
      cpu: 5
      memory: 12Gi
    requests:
      cpu: 5
      memory: 12Gi

zeebe-gateway:
  replicas: 1
  logLevel: debug
  env:
    - name: ZEEBE_LOG_APPENDER
      value: Stackdriver
    - name: ZEEBE_LOG_STACKDRIVER_SERVICENAME
      value: zeebe-gateway
    - name: ZEEBE_LOG_STACKDRIVER_SERVICEVERSION
      valueFrom:
        fieldRef:
          fieldPath: metadata.namespace
    - name: ATOMIX_LOG_LEVEL
      value: INFO
    - name: ZEEBE_GATEWAY_MONITORING_ENABLED
      value: "true"
    - name: ZEEBE_GATEWAY_THREADS_MANAGEMENTTHREADS
      value: "4"

tasklist:
  enabled: false

operate:
  enabled: false

As an example migration you can check this commit for the zeebe benchmarks.