option to skip slow-start in final bandwidth report
Closed this issue · 5 comments
GoogleCodeExporter commented
It would be useful to be able report the 'sustained' throughput after TCP
finishes ramping up and has stabilized.
Simple suggestion:
add a new flag: --skip N : ignore the first N seconds in the final report:
(default = 1)
Original issue reported on code.google.com by bltier...@es.net
on 6 May 2013 at 2:35
GoogleCodeExporter commented
could you work on this one next? It would be very useful to me to have soon.
Original comment by bltier...@es.net
on 27 Jun 2013 at 3:32
GoogleCodeExporter commented
You bet. I'll change the milestone to 3.0-Release.
When you say default = 1, do you mean it should always ignore the first second
of data unless there's an explicit --skip 0?
And let's see, what can we use for the one-letter flag... -S is reserved for
TOS. How about -K?
Original comment by jef.posk...@gmail.com
on 27 Jun 2013 at 3:42
- Added labels: Milestone-3.0-Release
- Removed labels: Milestone-3.1a1
GoogleCodeExporter commented
> When you say default = 1, do you mean it should always ignore the first
second of data unless there's an explicit --skip 0?
correct.
> And let's see, what can we use for the one-letter flag... -S is reserved for
TOS. How about -K?
I dont have a strong opinion, but maybe -I (ignore)
Original comment by bltier...@es.net
on 27 Jun 2013 at 3:47
GoogleCodeExporter commented
-I is good. Can I also change the long name to --ignore?
And how about making the ignore period in addition to the requested test
duration, instead of sliced out of it? So for example if the user requested a
ten second test, there would be one second of ramp-up and then ten seconds of
actual test, for a total of eleven seconds of runtime.
Original comment by jef.posk...@gmail.com
on 27 Jun 2013 at 3:51
GoogleCodeExporter commented
Original comment by jef.posk...@gmail.com
on 3 Jul 2013 at 7:03
- Changed state: Fixed