riak_core and crashed FSM vnode

Mike Oxford moxford at gmail.com
Wed Apr 6 19:41:09 EDT 2011

When you crash a riak_cord vnode with a console attached, what is the
expected behaviour (ba-dum, dump)?

I ran a riak_core vnode with some bad code (1/0) to cause a bad arith crash
in the FSM, initiated from the a console (via node "attach").

It crashed as expected, obviously, "** State machine <0.85.0> terminating"
The attached console was then hung.  I could ^C and break out of it...but I
chose ^D to close the connection.  Okay, now reattach.

I would expect that the vnode_master supervisor would have restarted the
FSM, as it's set to start_link in the supervisor's init.

However, when I reattached it was in a "dead" state.  I could not execute
any command, including "2+2."
^C-q to quit out, reattach, node is down (obviously.)

Is this "hang status" because I took it down while a shell was attached, or
is there something else which may prevent the vnode_master from
restarted the crashed vnode/FSM?


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.basho.com/pipermail/riak-users_lists.basho.com/attachments/20110406/305babf0/attachment.html>

More information about the riak-users mailing list