Operate throws "connection reset" when trying to connect to elasticsearch

Camunda 8 Topics
1

Hi,

  1. I tried to deploy Camunda 8 Self-Managed using Helm Chart into our Kubernetes environment.
  2. But, operate pod remains in error with “init: 0/2”. I think it’s failing in spec.initContainers{migration}.
  3. The warning message in the migration log for the operate pod is:
    io.camunda.operate.util.RetryOperation - Retry Operation Connect to Elasticsearch cluster [elasticsearch] at [http://marketplace-camunda-poc-test-elasticsearch:9200](http://marketplace-camunda-poc-test-elasticsearch:9200/) failed: java.util.concurrent.ExecutionException: java.net.SocketException: Connection reset org.elasticsearch.ElasticsearchException: java.util.concurrent.ExecutionException: java.net.SocketException: Connection reset at org.elasticsearch.client.RestHighLevelClient.performClientRequest(RestHighLevelClient.java:2695) ~[elasticsearch-rest-high-level-client-7.17.26.jar:8.13.4]
2

Hi @yooslee, welcome to the forums! There’s not enough information to go on here …

  • do you see errors in Tasklist, or does it start normally?
  • do you see errors in the Elastic pod?
  • can you share your values.yaml file (with sensitive info redacted)?
  • has it ever worked or was this the first attempt at a deployment?