search

datahub-project / datahub

The Metadata Platform for your Data Stack
https://datahubproject.io
Apache License 2.0
9.88k stars 2.93k forks source link

Helm Install under minikube causes high all CPU usage and Cluster goes down, killing elasticsearch as workarround #7371

Closed sapcode closed 1 year ago

sapcode commented 1 year ago

Describe the bug Helm Install under minikube causes high all CPU usage and Cluster goes down, killing elasticsearch as workarround.

To Reproduce Steps to reproduce the behavior:

  1. Install Minikube
  2. Install Current datahub using with following adjusted helm chart values.yaml: elasticsearch: enabled: true replicas: 1 minimumMasterNodes: 1 antiAffinity: "soft" fullnameOverride: "elasticsearch" imagePullPolicy: "Always" esJavaOpts: "-Xmx384m -Xms384m" maxUnavailable: 0 clusterHealthCheckParams: "wait_for_status=yellow&timeout=5s"

Expected behavior Installation should succeed

Screenshots During Helm install all CPU's of the cluster are used and cluster goes down sporadically: image

The helm installer aborts with following error: image

Desktop (please complete the following information):

Additional context When we kill the elasticsearch POD, the cluster stabilizes for a short while the Datahub system will boot up but within 1 hour sporadically without no queries in Datahub the all CPU high usage is appearing again bringing down the cluster: image

sapcode commented 1 year ago

Sorry wrong project... Issue is related to Openmetadata -> https://github.com/open-metadata/OpenMetadata/issues/10231