[Bro] Bro cluster requirements and manager logging backlog bug

[Hovsep Levi]

I can see a number of knobs that could make it happen but I don't know how
to go about scripting it.  I think it would:

- disable logging to manager (done automatically by having a logger node)
- bypass the single logger limit
- configure each logger to have a writer::kafka
- disable other writers if necessary
- check if the local worker is part of the same node for the local logger
(based on IP address I guess) and use that as a filter for the
worker2logger events

Starting from bro_init () I don't know how to do this or if it can be done
in conjunction with node.cfg or a custom-layout.bro.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mailman.ICSI.Berkeley.EDU/pipermail/bro/attachments/20161222/0b51920b/attachment.html