Zeek 3.1 introduced a first, experimental version of the new
supervisor framework that we expect to eventually replace ZeekControl
as the primary mechanism to run Zeek clusters, both single- and
multi-system. See the earlier blog posting for more context:
https://blog.zeek.org/2019/03/beyond-brocontrol-new-process.html.
The Zeek docs summarize what's now available in 3.1:
https://docs.zeek.org/en/current/frameworks/supervisor.html. (Thanks,
to Jon for all that great work!)
I'd like to start planing out the next steps here, and hope that we
can start deprecating (*not* remove) ZeekControl with Zeek 4.0 late
this year. To that end, I drafted a supervisor roadmap for the next
couple of Zeek versions, with a proposal on which parts of
ZeekControl's functionality to port over into the new supervisor
world, and when:
https://docs.google.com/spreadsheets/d/1oNYqHMU1b6FEfST8F86bdpv9Bpl6M58qq4AZULCIYlU/edit?usp=sharing
I'd be interested in thoughts on that, both in terms of timeline and
generally any questions or concerns you might have. Best places for
feedback are this mailing list and the "Supervisor Project" ticket
(https://github.com/zeek/zeek/issues/582).
Robin
(I'll keep this on zeek-dev for now; will post more broadly as well
once the timeline stabilizes).