`crm config delete` should check or wait until pacemaker scheduler is IDLE

[zzhou1]

Here is the example,

1. setup a ocfs2 cluster
   adm@tws-1:~> sudo crm cluster init -ys /dev/disk/by-partlabel/sbd-tw -o /dev/disk/by-partlabel/ocfs2

2. sudo crm resource stop ocfs2-clone

3. sudo crm config del ocfs2-clone ocfs2-group ocfs2-clusterfs dlm

The improvement is about step#3, to let it by default to check or wait until pacemaker scheduler is IDLE before proceed. Otherwise, there could be the faulty situation step#2 RA does not fully stopped, the delete operations in step#3 can fail badly in the middle.