QoS and Jetlang

[John M]

Hi,

I'm fairly new here, so my apologies if these are common questions. I
did not see them mentioned in the group history, but am sure others
have pondered them.

We're considering how to manage a QoS with Jetlang, and are curious
about what others are doing to manage fiber priorities and inspecting
channels for backlog, etc.

Thanks,

[Mike Rettig]

In production systems, I provide my own BatchExecutor implementation
that monitors the queue depth and the execution time of all events.

http://code.google.com/p/jetlang/source/browse/trunk/src/main/java/org/jetlang/core/BatchExecutorImpl.java

This executes on the consuming thread. The custom implementation times
every execution using System.nanoTime() to detect slow events and
calculate avg latencies. The downside to this approach is that it
can't detect if a fiber has deadlocked or has never been started since
it only gets called when events are consumed on the fiber. The
executor also catches any and all exceptions and notifies the system
maintainers of the error.

Mike

> --
> You received this message because you are subscribed to the Google Groups "jetlang-dev" group.
> To post to this group, send email to jetla...@googlegroups.com.
> To unsubscribe from this group, send email to jetlang-dev...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/jetlang-dev?hl=en.
>
>