Log output control

Question

If you run node as a slave, all of its log output is given to the wizard. However, in my setup, I do not want to have a wizard, and I have nodes that automatically detect and join the cluster as they see fit. However, I would still like all of the cluster log output to be the only node syntax. Is there a way to dynamically log the node, as if it were running as a slave? Otherwise, do I need to change every error_handler set to redirect the output to where I want it to go?

Here would be my ideal setup: I switch the switch and all the nodes in the cluster send everything that happens in any of the tty node names - io: format or sasl or whatever you have - instead of a single node, where it maps to tty and goes into round robin files. What would make this a reality?

Answer 1

To do this, use group_leader. Have you checked this link?

Answer 2

I think http://jkvor.com/log-roller might be the answer, although it will not record io: format calls, I don't think so. However, if you restrict your logging functions to the error_logger module, it should work fine.