Hello, I’m trying to do docker-compose on datahub-...
# troubleshoot
w
Hello, I’m trying to do docker-compose on datahub-gms with the latest head version with the below configs. But getting below error, error message from docker logs:
Copy code
2023-02-14 20:25:50,405 [main] WARN  c.l.metadata.entity.EntityService:798 - Unable to produce legacy MAE, entity may not have legacy Snapshot schema.
java.lang.UnsupportedOperationException: Failed to find Typeref schema associated with Config-based Entity
Configs used for gms:
Copy code
datahub-gms:
              container_name: datahub-gms
              environment:
                    - DATAHUB_UPGRADE_HISTORY_KAFKA_CONSUMER_GROUP_ID=generic-duhe-consumer-job-client-gms
                    - EBEAN_DATASOURCE_USERNAME=xyz
                    - EBEAN_DATASOURCE_PASSWORD=xyz
                    - EBEAN_DATASOURCE_HOST=xyz
                    - EBEAN_DATASOURCE_URL=jdbc:<postgresql://xyz>
                    - EBEAN_DATASOURCE_DRIVER=org.postgresql.Driver
                    - KAFKA_BOOTSTRAP_SERVER=broker:29092
                    - KAFKA_SCHEMAREGISTRY_URL=<http://schema-registry:8081>
                    - ELASTICSEARCH_HOST=elasticsearch
                    - ELASTICSEARCH_PORT=9200
                    - ES_BULK_REFRESH_POLICY=WAIT_UNTIL
                    - ELASTICSEARCH_INDEX_BUILDER_SETTINGS_REINDEX=true
                    - ELASTICSEARCH_INDEX_BUILDER_MAPPINGS_REINDEX=true
                    - NEO4J_HOST=<http://neo4j:7474>
                    - NEO4J_URI=<bolt://neo4j>
                    - NEO4J_USERNAME=neo4j
                    - NEO4J_PASSWORD=datahub
                    - JAVA_OPTS=-Xms1g -Xmx1g
                    - GRAPH_SERVICE_DIFF_MODE_ENABLED=true
                    - GRAPH_SERVICE_IMPL=neo4j
                    - ENTITY_REGISTRY_CONFIG_PATH=/datahub/datahub-gms/resources/entity-registry.yml
                    - ENTITY_SERVICE_ENABLE_RETENTION=true
                    - MAE_CONSUMER_ENABLED=true
                    - MCE_CONSUMER_ENABLED=true
                    - PE_CONSUMER_ENABLED=true
                    - UI_INGESTION_ENABLED=true
                    - METADATA_SERVICE_AUTH_ENABLED=false
              hostname: datahub-gms
              image: ${DATAHUB_GMS_IMAGE:-linkedin/datahub-gms}:${DATAHUB_VERSION:-head}
              ports:
              - ${DATAHUB_MAPPED_GMS_PORT:-8080}:8080
Can anyone help me on this? Thanks
a
Hi @witty-actor-87329, there are some new environmental vars in the latest version that you may need to include
We’re working on getting notes out for those now
w
Thanks @astonishing-answer-96712 can you please point me to the environmental vars in the github path that I can refer to?
@astonishing-answer-96712 Are the below env vars that you are mentioning:
Copy code
- DATAHUB_SERVER_TYPE=${DATAHUB_SERVER_TYPE:-quickstart}
      - DATAHUB_TELEMETRY_ENABLED=${DATAHUB_TELEMETRY_ENABLED:-true}
I tried adding these but still getting the same error.
a
Hi @witty-actor-87329 please check this thread for context, but it lists the new vars: https://datahubspace.slack.com/archives/C02AD211493/p1676413846218329?thread_ts=1676395703.658869&amp;cid=C02AD211493
w
@astonishing-answer-96712 Still getting the same error after adding the new env vars
a
just to confirm, you are on 0.10.0?
w
We were on 0.8.0 and trying to upgrade it to 0.10.0
During upgrading datahub-gms container, getting this error
a
CC: @orange-night-91387 This is likely due to the upgrade container hanging, correct?
o
The message in the first thread is an ignorable warning message that just is saying that the legacy mae is unable to be produced for an entity that is not defined with the legacy snapshot schema. That's a non issue. What is the exact behavior you are seeing? This is the last message appears in the logs and GMS is hanging on start-up? How are you deploying DataHub? There is an additional component in the latest version that needs to run and update the indices with new mappings before GMS will start.
If you are using a custom compose file and not the one we supply you will need to update it
f
Hi @witty-actor-87329 . Did you find a solution? We are facing the same error message when upgrading from 0.9.6.1 to 0.10.0 (using Helm). Interestingly, the first pod started successfully, but the further pods can not start due to this issue.
Working fix for me: Redeploy v0.10.0 again using Helm. Error is no longer there.