: Clean-ups
cover : Clean-ups
Quoting Arnaldo Carvalho de Melo:
| most things merged, except for the packet size and | TRFC equation patches, that I thought a day or two would be nice for | further discussion.
I have tidied up what remained and uploaded the up-to-date remaining patches on to
http://www.erg.abdn.ac.uk/users/gerrit/dccp/patch-backlog/
There are two small patches which I send as RFC, for the first I need a consensus:
Patch 1: Removes a typo in setting the nofeedback timer.
=> Question: The nofeedback timer is currently set to time out after
max(t_rto, 2 * s/X) = max(max(4*R, 1 second), 2 *s/X)
This is non-standard and has been flagged in the code as such
I have conferred with a colleague who does a lot of simulations in this field and he says doing this is bad, especially when the RTTs are small. Therefore, since it is both non-standard and obviously not a good idea, I want to suggest revising the entire patch so that the nofeedback timer expires after max(4*R, 2 * s/X) as per RFC 3448.
OK with this suggestion?
Patch 2: Implements a useful idea which Ian suggested while discussing one of the
previous patches: if the len' arguments of ccid_{send_packet,packet_sent}() are unsigned, then we do not need to perform extra checks against len > 0' The patch implements this idea and, as a result, another unused block/test in ccid3.c can go.
Gerrit
Quoted 5 lines
Patch 1: Removes a typo in setting the nofeedback timer.
=> Question: The nofeedback timer is currently set to time out after
max(t_rto, 2 * s/X)=max(max(4*R, 1 second), 2 *s/X)
Quoted 8 lines
This is non-standard and has been flagged in the code as such
I have conferred with a colleague who does a lot of simulations in this field and he says doing this is bad, especially when the RTTs are small. Therefore, since it is both non-standard and obviously not a good idea, I want to suggest revising the entire patch so that the nofeedback timer expires aftermax(4*R, 2 * s/X)as per RFC 3448.
OK with this suggestion?
The code here predates me but I think the reason for it might be to do with RFC4342 (remember to read 3448 in conjunction with this).
If the sender never receives a feedback packet from the receiver, and as a consequence never gets to set the allowed sending rate to one packet per RTT, then the sending rate is left at its initial rate of one packet per second, with the nofeedback timer expiring after two seconds. The allowed sending rate is halved each time the nofeedback timer expires. Thus, if no feedback is received from the receiver, the allowed sending rate is never above one packet per second and is quickly reduced below one packet per second.
NB I'm not saying the old code is correct - I'm just guessing that the above is reason for it... Your thoughts Gerrit? (I haven't re-read the code or checked any of it out properly)
Ian
Signature · 7 lines
Ian McDonald Web: http://wand.net.nz/~iam4 Blog: http://imcdnzl.blogspot.com WAND Network Research Group Department of Computer Science University of Waikato New Zealand
Quoted 12 lines
The code here predates me but I think the reason for it might be to do with RFC4342 (remember to read 3448 in conjunction with this).
If the sender never receives a feedback packet from the receiver, and as a consequence never gets to set the allowed sending rate to one packet per RTT, then the sending rate is left at its initial rate of one packet per second, with the nofeedback timer expiring after two seconds. The allowed sending rate is halved each time the nofeedback timer expires. Thus, if no feedback is received from the receiver, the allowed sending rate is never above one packet per second and is quickly reduced below one packet per second.
Quoted 4 lines
NB I'm not saying the old code is correct - I'm just guessing that the above is reason for it... Your thoughts Gerrit? (I haven't re-read the code or checked any of it out properly)
Just realised I didn't explain what I actually meant. I'm wondering whether initially no feedback timer was used to maintain 1 packet per second (which is the wrong way to do it).
Signature · 7 lines
Ian McDonald Web: http://wand.net.nz/~iam4 Blog: http://imcdnzl.blogspot.com WAND Network Research Group Department of Computer Science University of Waikato New Zealand
Quoting Ian McDonald:
| > The code here predates me but I think the reason for it might be to do | > with RFC4342 (remember to read 3448 in conjunction with this). | > | > If the sender never receives a feedback packet from the receiver, and | > as a consequence never gets to set the allowed sending rate to one | > packet per RTT, then the sending rate is left at its initial rate of | > one packet per second, with the nofeedback timer expiring after two | > seconds. The allowed sending rate is halved each time the nofeedback | > timer expires. Thus, if no feedback is received from the receiver, | > the allowed sending rate is never above one packet per second and is | > quickly reduced below one packet per second. | > | > NB I'm not saying the old code is correct - I'm just guessing that the | > above is reason for it... Your thoughts Gerrit? (I haven't re-read the | > code or checked any of it out properly) | > | Just realised I didn't explain what I actually meant. I'm wondering | whether initially no feedback timer was used to maintain 1 packet per | second (which is the wrong way to do it).
I checked RFC 4342 again - there is no news with regard to RFC 3448. And in RFC 3448 it is in the way suggested - use max(4*R, 2*s/X). I think it is better to use this value, as described.
There is a lot of cruft from the old code. I would like to go over that - I have been studying older FreeBSD patches and now see the relationships.
Quoted 20 lines · Gerrit Renker
Quoting Ian McDonald:
| > The code here predates me but I think the reason for it might be to do | > with RFC4342 (remember to read 3448 in conjunction with this). | > | > If the sender never receives a feedback packet from the receiver, and | > as a consequence never gets to set the allowed sending rate to one | > packet per RTT, then the sending rate is left at its initial rate of | > one packet per second, with the nofeedback timer expiring after two | > seconds. The allowed sending rate is halved each time the nofeedback | > timer expires. Thus, if no feedback is received from the receiver, | > the allowed sending rate is never above one packet per second and is | > quickly reduced below one packet per second. | > | > NB I'm not saying the old code is correct - I'm just guessing that the | > above is reason for it... Your thoughts Gerrit? (I haven't re-read the | > code or checked any of it out properly) | > | Just realised I didn't explain what I actually meant. I'm wondering | whether initially no feedback timer was used to maintain 1 packet per | second (which is the wrong way to do it).
Quoted 7 lines
I checked RFC 4342 again - there is no news with regard to RFC 3448. And in RFC 3448 it is in the way suggested - use
max(4*R, 2*s/X). I think it is better to use this value, as described.There is a lot of cruft from the old code. I would like to go over that - I have been studying older FreeBSD patches and now see the relationships.
OK. Go ahead. I will ack patch.
The reason why I'm just acking patches still is haven't had good chance to test properly yet.
I'm also still thinking about parameter s pass in and will still review tfrc equation stuff.
Regards,
Ian
Signature · 7 lines
Ian McDonald Web: http://wand.net.nz/~iam4 Blog: http://imcdnzl.blogspot.com WAND Network Research Group Department of Computer Science University of Waikato New Zealand