Hi Stefan,

  

Your suggestions have no effect for me. Still errno 16. I think at the 
moment it is not possible to connect concurrently via l2cap if one 
device is down. But as a workaround I can use non-blocking connect (but 
with a timeout of 60 seconds just to prevent deadlocks) and simply 
reduce the pageto. My connection attempts are sequential anyway because 
of the master-slave-limit. But I would suggest to change the behavior of 
l2cap to allow more concurrent connection attempts even if some devices 
are down.
    

use hcidump and check the result code of the failing HCI connection
complete event. Look them up in the Bluetooth specification. This may
lights this behaviour a little bit more up.