On Thu, Jul 30, 2015 at 11:52 AM, Sean DuBois <s...@siobud.com> wrote:
> I ran into this same issue and ended up putting a little REST server on > top of fail2ban (using the socket that fail2ban-client uses) and if > something is banned on one server hit the API of all its siblings. > > So, each fail2ban instance would send lots of requests, one per sibling, right? In my case, I am adding a process to which clients connect as subscribers, and where fail2ban-cluster (adhoc process that reads fail2ban.log), monitors for ban/unbans and sends one message to the distribution server, which then all subscribers receive. I think it is quite more fluid. Of course, it was also an excuse to start working with this kind of message queuing systems. Check out http://zeromq.org/
------------------------------------------------------------------------------
_______________________________________________ Fail2ban-users mailing list Fail2ban-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/fail2ban-users