IETF 70 BLISS WG meeting

1
IETF 70 BLISS WG meeting

• Automated Handling
• draft-elwell-bliss-ach-analysis-00draft-elwell-bl
  iss-dnd-01

2
Topics

• draft-elwell-bliss-ach-analysis (automatic call
  handling analysis)
• draft-elwell-bliss-dnd-01 (do not disturb)
• Discussion

3
draft-elwell-bliss-ach-analysis

• Identifies many examples of conditions that can
  give rise to ACH
• Identifies various actions that can be taken when
  conditions are met (typically forward elsewhere
  or reject)
• Discusses a number of problem areas
• Identifies questions that BLISS needs to answer
  to take this topic further

4
ACH problem areas

• Interactions between proxy and UA
• Which wins depends on whether call ever reaches
  the UA
• Also depends whether proxy can figure out what UA
  has done and override with own action
• Conflict between UAs
• Depends whether serial or parallel forking
  (latter less deterministic)
• Proxy determination of what has happened at UA
• Are response codes precise enough (e.g., for
  DND)?
• Informing the calling user
• Similar to above, but privacy may play a part
• Redirection versus retargeting

5
ACH problem areas (continued)

• Scope of conditions
• Does a proxy wait for all UAs to respond, or act
  when first UA reports a given condition?
• Scope of 6xx response
• Does it include voicemail?
• Does it include forwarding to a different AoR?
• Configuring the proxy
• Several different ways (e.g., web page, web
  service, CSTA, CPL upload)
• Related issue of whether UA and user can view
  what is configured at proxy

6
ACH - other points discussed

• Proxy versus UA
• Lots of good reasons for doing ACH at the proxy
• But can be particular reasons for UA-specific ACH
• May be different considerations for enterprise
  and residential environments
• Avoiding inconsistent configurations
• Use of configuration framework to ensure
  consistency, whilst still allowing UA-specific
  behaviour where required
• Possible approach is to download policy to UA
  indicating whether proxy is responsible for ACH

7
ACH questions identified

1. Do we need to make any recommendations on
   proxy-based versus UA-based ACH?
2. Is provisioning (e.g., using SIP config
   framework) a suitable solution for ensuring
   consistency between UAs and between UAs and
   proxy?
3. Do we need to make any recommendations on
   response codes and other information in responses
   between UAS and proxy?

8
ACH questions identified (continued)

1. Do we need to make any recommendations concerning
   meaning and use of 6xx response codes?
2. Do we need to specify a recommended or default
   method for a UA or PDS to configure its proxy and
   discover how its proxy is configured?
3. To what extent is it necessary to make
   recommendations concerning keeping the calling UA
   informed of what has happened?

9
draft-elwell-bliss-dnd-01

• Changes from 00
• additional means of invoking DND (casual, based
  on sensing a user's current activity)
• distinction between conveying DND to a proxy and
  conveying the condition to the caller
• problem of indicating DND to a caller when the
  condition does not relate to the original called
  party (e.g., the call was forwarded first)
• whether to wait for all branches to respond or
  act immediately a branch reports DND
• Further work will depend on general ACH work (but
  not a total overlap, e.g., presence aspects of
  DND)