As reported on the mailing list thread http://lists.slony.info/pipermail/slony1-general/2014-May/012811.html Slonik can hang during a failover because a non failover candidate node can be using a second failed node as its subscription provider. It should be reconfigured to use one of the remaining nodes as the subscription source but this isn't happening.
Created attachment 202 [details] proposed patch
(In reply to comment #1) > Created an attachment (id=202) [details] > proposed patch Tested here and behaves as expected. Thanks again Steve.
This was committed http://git.postgresql.org/gitweb/?p=slony1-engine.git;a=commit;h=f97ef5aed23b031da586ef6e46f9875c5d99b005