Metric for requests queued up and eventually times out?

Hi all, consider the following setup:

  • linkerd has started, and created all the listeners
  • Since the Consul DNS is bogus, it is not able to get any downstream client
  • Incoming requests timeout after 10 seconds with an error similar to:
    Internal error processing ping: 'com.twitter.finagle.RequestTimeoutException: exceeded 10.seconds to while dyn binding /svc/ping. Remote Info: Not Available'"

Is there a metric that we can use to track this queue size?

In addition, what configurations are relevant here? Basically, can I not wait for 10 seconds if there is no backend?


There is no queue but you can configure the timeout with the bindingTimeoutMs property:

Thanks @Alex!

What does “binding” mean here in this case ? Is it - I got a request, i need to find an appropriate downstream to forward it to?

Yes, that is correct.

1 Like

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.