MDC Troubleshooting Checklist

Has each node in the source/sink cluster(s) been tuned?

Network connectivity verified

• iperf etc
• Using EC2? Suggest this doc: http://docs.basho.com/riak/latest/ops/tuning/aws/

Static data

• priority (urgent/high/low)
• version of Riak EE
  • from sources AND sinks
• version of replication
• topology
  • how many clusters
  • bidirectional?
• app.config config
  • from sources AND sinks

Command line status

• colllected from the cluster_leaders on source AND sink
  • riak-repl status
  • riak-repl connections
  • riak-admin status

Erlang state

• cluster_mgr state rp(sys:get_status(riak_core_cluster_manager)).
• connection_mgr state rp(sys:get_status(riak_core_connection_manager)).
• service_mgr state rp(sys:get_status(riak_core_service_manager)).
• riak_repl2_leader state rp(sys:get_status(riak_repl2_leader_gs)).
• riak_repl_leader_gs state rp(sys:get_status(riak_repl_leader_gs)).

Fullsync issues

• riak_repl2_fscoordinator rp(sys:get_status(riak_repl2_fscoordinator)).
• riak_repl2_fscoordinator_serv rp(sys:get_status(riak_repl2_fscoordinator_serv)).

Realtime issues

• riak_repl2_rtq rp(sys:get_status(riak_repl2_rtq))