kafka-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Ewen Cheslack-Postava <e...@confluent.io>
Subject Re: ConnectStandalone with no starting connector properties
Date Mon, 05 Dec 2016 04:50:13 GMT

Sure, we'd be happy to commit a patch that removes this restriction.

In practice, for most folks its a bit simpler to generate the config file
in whatever deployment system they are using than to start the standalone
server separately and make an HTTP request to trigger creation of the
connector (which needs to happen on ever restart of the process, even those
that might not be managed by your deployment system, e.g. due to unexpected
process failure that is recovered by something like your init system).
However, I don't see any reason we can't support the approach it sounds
like you're proposing.


On Fri, Dec 2, 2016 at 10:42 AM, Micah Whitacre <mkwhitacre@gmail.com>

> I'm curious if there was an intentional reason that Kafka Connect
> standalone requires a connector properties on startup?[1]
>  ConnectDistributed only requires the worker properties.  ConnectStandalone
> however requires the worker properties and at least one connector
> properties.  Since connectors can still be managed for standalone using the
> REST API, I was wondering if that restriction could be lifted?  If so I'll
> log an enhancement JIRA but wanted to make sure I wasn't missing something
> obvious.
> Thanks,
> Micah
> [1] -
> https://github.com/apache/kafka/blob/93804d50ffc40fb6cdc61c073f4a62
> f0931f042d/connect/runtime/src/main/java/org/apache/kafka/connect/cli/
> ConnectStandalone.java#L60-L63


  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message