[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: NM WG call this Wednesday, 8am PDT
On Tuesday, May 13, 2003, at 11:10 AM, Cottrell, Les wrote:
A. For path.reordering.*, shouldn't there be an indicator of the
separation of the test packets. One will see much lower re-ordering
for say ping packets separated by 1 second, versus say reordering of
IP packets sent by TCP, or test packets sent with say 20 usec
separation.
you are right. In
http://www.ietf.org/internet-drafts/draft-ietf-ippm-reordering-02.txt,
they seem to solve this by including the start and stop time for the
test. How do we want to handle this?
B. The use of Mbits/s with a unit32 for various values (e.g.
path.reordering.* and path.bandwidth.* would seem to preclude using
this for slower speed links, e.g. for wireless, or for links to
developing nations, or modems/DSL etc.
yep. So we should use real32, right?
In the last sentence of para 2: I am unclear why hop-by-hop
measurement of available bandwidth provides prediction of the
performance available from each source to a destination. This is not
to say it is not useful to know the hop-by-hop characteristics.
If you have the option of copying a file from 2 sources (eg: half of
the file from each source), you only want to do this if the 2 paths do
not share the bottleneck hop. hop-by-hop info is needed for to
determine this.
------------------------------------------------------------------------
-------------------
Brian L. Tierney, Lawrence Berkeley National Laboratory (LBNL)
1 Cyclotron Rd. MS: 50B-2239, Berkeley, CA 94720
tel: 510-486-7381 fax: 510-495-2998 efax: 240-332-4065
bltierney@lbl.gov http://www-didc.lbl.gov/~tierney
------------------------------------------------------------------------
------------------