Re: [manet-dlep-rg] Slight issue with Data items

On Mar 10, 2014, at 12:40 PM, Rick Taylor <rick@tropicalstormsoftware.com> wrote:

> Hi All,
> 
> I've been ruminating over Data Items and their default handling as we discussed in London, i.e. an implementation MUST silently drop any unrecognised Data Items it receives.
> 
> But, there is some inconsistency over what a Data Item is (in my mind), and so I ask:  Is a Data Item *any* signal payload, or is it one of the payload items in a Destination Up/Down/Update message?
> 
> The reason I ask is that there is a class of TLVs in DLEP that appear in session initiation (and possibly discovery) that are semantically different from 'regular' TLVs in session signals: Version information, TCP server addresses, etc...  Now, are these TLVs treated as Data Items as far as the handling rules apply?  What happens if you receive a Version Data Item in a Destination_Update:  The implementation should recognise it, but not expect it where it finds it.
> 
> So, have I revealed a new class of Data Item that needs it's own rules and status code, or do the existing rules still work?

I don't know if you've revealed a new class of data item, but you've opened an area for discussion that we haven't talked about yet - Data Items that are understood, but not meaningful for the DLEP signal. Version information in a destination update is a perfect example of that.

I'll offer up a slight modification of the "London Rule" :  Experimental Data Items (and Vendor-specific TLVs) that are not supported by an implementation should be parsed (e.g. "stepped over') and silently dropped. Data Item TLVs that are not *understood* (e.g. TLV number not registered, or "reserved"), OR are out-of-context (like a Version TLV in a destination update message) should cause PEER_TERMINATION and closure of TCP session.

Regards,
Stan

> 
> In summary, do we need to define "Status Code <N>: Data Item not allowed in last signal"
> 
> Hmm... something for you all to ponder at the start of the week...
> 
> Rick
> 
> P.S> Is everyone happy for me to post the summary report to the WG mailing list?  If I hear no objections by close of play Tuesday, I'll just forward it.
> _______________________________________________
> manet-dlep-rg mailing list
> manet-dlep-rg@ietf.org
> https://www.ietf.org/mailman/listinfo/manet-dlep-rg