[Iccrg] SSDT Scope - summary

[Mark Allman]

> Personally, I'm not a fan of QuickStart/JumpStart, because it doesn't
> give other flows enough time to back off.  

I think this sentence lumps things together that should not be lumped
together.  QuickStart should not have the problem of interacting with
other flows because each router should be tracking the capacity being
consumed and so should not agree to a QS request that would in fact
cause interference with competing traffic.  (Certainly there are cases
that can cause mistakes---but, these are just that: mistakes.)

The consideration for JumpStart interfering with other traffic is
completely justified because JS is at the complete opposite end of the
spectrum from QS.  While QS tries to be quite careful about checking for
a proper rate, JS crudely simply starts at whatever rate the endpoint
feels like.  Clearly this latter approach is much more likely to
interfere with competing traffic.

allman



-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 185 bytes
Desc: not available
Url : http://oakham.cs.ucl.ac.uk/pipermail/iccrg/attachments/20080416/59f89856/attachment.bin