TODO

NSIv2 BOD Service
 - Check nsidev logs and improve them
 - Better error messages on bindings (sigh...)




##
 Better debugging on SSH backends

Current issues:

- Make replyTo optional (assuming sync. query)

- Better aggregate error handling
- Better logging in aggregation failures
- NSI2 Detail Query

- Label support in topology and backends
- Rework all error throwing handling to new error infrastructure
- Rename shared.resource to shared.soapresource

Registry, they are not necessarely events or service. Just a registry. (just some renaming)

Log rotation

Long term:
  Documentation
  Persistence for connections
  Mailing list
  Easy way of seeing connections in OpenNSA (basic web interface is there)
  Deal with bandwidth in calendar (requires joint resource concept)
  Ability to generate topology from NRM mapping / internal topology description

  Think about AAI and Ports, Probably want private ports
   - How do we represent port access?


General:
  Seperate protocol stack from connection (done for provider, client still needs work)
  Profiling of protocol stack / backends
  Better connection names in client - use UUIDs for global ids
  Make RPC + callback timeout configurable
  Implementation of AutoGOLE topology distribution scheme
  Client should be able to breakout request itself
  Security params clean up (see mail from Jeroen)


Backends:
  NCS backend
  Force10: Should do shutdown when interfaces are not in use
  Test on Juniper EX switch
  Test on Juniper virtual router
  JunOS:
    Failure handling is terrible for command sending
    Log success/failure instead of commands (only on debug)
  SSH module does not notifiy login timeouts