Logstash and multiple outputs: how can I circumvent unreliable outputs?


in a test scenario (still using Logstash 1.5) we like to push logevents to two separate outputs.
The first one is 'elasticsearch' the second one shall be 'redis'.

Right now we experience errors in the whole Logstash instance when one output is not available.

Are these blocking issues removed in newer versions of Logstash or is this a general weakness of the design?

Is there a way to circumvent this?
Like having an output that drops events when the endpoint is not available?

This behaviour is intentional as LS, there is no way around it.

Hi @warkolm,

thank you very much for your answer.

I was kind of certain before this would be the answer. And it is. :disappointed:

I'll have to find some different way to deliver these events.