Problem Statement
When using Redpanda Connectors in self-managed environments using helm-chart versions 5.8.3 and higher there may be no configured connectors displayed in Console>Connectors.
This is due to a bug in that the timeout settings for Console:Connectors are set to a zero value and causing Console to immediately timeout when calling the endpoint to return the list of managed connectors.
Detail
Errors seen in Console logs :
{"level":"warn","ts":"2024-09-26T10:42:39.876Z","msg":"failed to list connectors from Kafka connect cluster","cluster_name":"connectors","cluster_address":"http://redpanda-connectors.server.svc.cluster.local:8083","error":"Get \"http://redpanda-connectors.server.svc.cluster.local:8083/connectors?expand=info&expand=status\": context deadline exceeded"}
This is happening due to Console having these timeout parameters for the connect configuration:
connect:
connectTimeout: 0
readTimeout: 0
requestTimeout: 0
Workaround/Resolution
Bug #1540 was created to address this issue and this document will be updated when available in a release
To workaround the issue :
set the following for the connect timeouts:
connect.connectTimeout=15s
connect.readTimeout=60s
connect.requestTimeout=6s
helm-chart
set the timeout settings using values/set.
For example with --set:
helm install redpanda redpanda/redpanda \
--version 5.9.5 \
--namespace redpanda \
--create-namespace \
--set external.domain=customredpandadomain.local \
--set statefulset.initContainers.setDataDirOwnership.enabled=true \
--set connectors.enabled=true \
--set console.config.connect.connectTimeout=15s \
--set console.config.connect.readTimeout=60s \
--set console.config.connect.requestTimeout=6s
helm-chart/operator
Set the connect timeouts
connect.connectTimeout=15s
connect.readTimeout=60s
connect.requestTimeout=6s
as part of : spec.clusterSpec.console.console.config