[net2-wg] CTP testing status

[Rodrigo Fonseca]

Hi Om,

try increasing the size of the DebugMessagePool and DebugMessageQueue.
Of course no queue will help if the aggregate rate of messages is larger
than the serial bandwidth. What is your estimate of the rate?

If this continues to be a problem, we can try to look into the stat
variables inside of UARTDebugSenderP.nc . We haven't done that, but one way
would be to have the node periodically send a message with the values of the
stats it collects.

Rodrigo


On Wed, Aug 27, 2008 at 2:10 PM, Omprakash Gnawali <gnawali at usc.edu> wrote:

> I haven't noticed something this serious on non-root nodes.
>
> - om_p
>
> On Wed, Aug 27, 2008 at 2:08 PM, Razvan Musaloiu-E. <razvanm at cs.jhu.edu>
> wrote:
> > Hi!
> >
> > Is happening even for nodes that don't relay too much traffic?
> >
> > --
> > Razvan ME
> >
> > On Wed, 27 Aug 2008, Omprakash Gnawali wrote:
> >
> >> Looking at the logs, I noticed something strange. On this file:
> >> http://enl.usc.edu/~om_p/net2/ctp-cleanup/run6/log0<http://enl.usc.edu/%7Eom_p/net2/ctp-cleanup/run6/log0>
> >>
> >> On line 25010, the log seq number (the last two bytes) is 0x3544.
> >> On line 25012, the next log entry, the sequence number is 0x3944.
> >>
> >> This is a big jump!
> >>
> >> Similar, there is a jump from 0x3e70 (line 27640) to 0x40b9 (line
> 27642).
> >>
> >> There are many other similar jumps in that log.
> >>
> >> Many sequence numbers just not there:
> >>
> >> $ grep SENT_MSG txtlog0
> >> ...
> >> 4622: NET_C_FE_SENT_MSG 0 246 0 34645 0
> >> 4634: NET_C_FE_SENT_MSG 0 247 0 34734 0
> >> 4643: NET_C_FE_SENT_MSG 0 248 0 34792 0
> >> 4663: NET_C_FE_SENT_MSG 0 249 0 34872 0
> >> 4671: NET_C_FE_SENT_MSG 0 7 0 35824 0
> >> 4681: NET_C_FE_SENT_MSG 0 8 0 35871 0
> >> 4699: NET_C_FE_SENT_MSG 0 9 0 35941 0
> >> 4718: NET_C_FE_SENT_MSG 0 10 0 36006 0
> >> 4762: NET_C_FE_SENT_MSG 0 11 0 36079 0
> >> 4791: NET_C_FE_SENT_MSG 0 12 0 36168 0
> >> ...
> >>
> >>
> >> - om_p
> >>
> >> On Mon, Aug 25, 2008 at 8:15 AM, Omprakash Gnawali <gnawali at usc.edu>
> >> wrote:
> >>>
> >>> I looked at the data from my experiments. I am getting strange result.
> >>> For up to 1000-1400s, I get close to 100% pkts at the root then it
> >>> plummets to a long term average of 50-70%. This is with or without the
> >>> change that triggers route update if current etx is larger than the
> >>> min etx by 20. With that change, the result is better than without but
> >>> we are getting pretty poor performance. I am trying to understand if
> >>> this has to do with my instrumentation or CTP.
> >>>
> >>> - om_p
> >>>
> >> _______________________________________________
> >> net2-wg mailing list
> >> net2-wg at millennium.berkeley.edu
> >> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/net2-wg
> >>
> >
> _______________________________________________
> net2-wg mailing list
> net2-wg at millennium.berkeley.edu
> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/net2-wg
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: https://www.millennium.berkeley.edu/pipermail/net2-wg/attachments/20080827/d66c10f1/attachment.htm