Prue, W., and Postel, J., "Something a Host Could Do with Source Quench", Request for Comments (RFC) 1016, July 1987.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....the algorithm follows immediately: The multiplicative decrease remains a multiplicative decrease (e.g. double the interval between packets) But subtracting a constant amount from interval does not result in an additive increase in bandwidth. This approach has been tried, e.g. 18] and [24], and appears to oscillate badly. To see why, note that for an inter packet interval and decrement , the bandwidth change of a decrease interval by constant policy is 1 1 a non linear, and destablizing, increase. An update policy that does result in a linear increase of bandwidth over time ....

PRUE,W.,AND POSTEL,J. Something A Host Could Do with Source Quench.ARPANET Working Group Requests for Comment, DDN Network Information Center, SRI International, Menlo Park, CA, July 1987. RFC-1016.

....in a core TRIC router is proportional to the number of marked packets in the output queue that have no corresponding ACK in the backward queue. This remains to be further investigated. 5.2.1. Source quench An alternative stateless mechanism is to allow core nodes generate a Source Quench [10] message towards the source, which gets intercepted by the ingress node. This has the slight overhead of generating new control messages, which again are limited in a probabilistic way by using RED algorithms. Evidently, if we would want to apply TRIC, the edge node would have to wait for an ACK ....

Prue, W., Postel, J., "Something a host could do with source quench", RFC 1016, July 1987.

....the receipt of a single ECN message serves as a notification of congestion to the TCP source. At the same time, the guidelines ensure that the TCP source does not respond to ECN messages more frequently than necessary. Guidelines: TCP s response to ECN should be similar, over 1 As an example, [PP87] suggested that the gateway send Source Quench messages when the queue size exceeds a certain threshold. longer time scales, to its response to a dropped packet as an indication of congestion. Over smaller time scales (e.g. one or two round trip times) TCP s response to ECN can be less ....

Prue, W., and Postel, J., "Something a Host Could Do with Source Quench", Request for Comments (RFC) 1016, July 1987.

....the algorithm follows immediately: The multiplicative decrease remains a multiplicative decrease (e.g. double the interval between packets) But subtracting a constant amount from interval does not result in an additive increase in bandwidth. This approach has been tried, e.g. 19] and [24], and appears to oscillate badly. To see why, note that for an inter packet interval I and decrement c, the bandwidth change of a decrease interval by constant policy is 1 I # 1 I c a non linear, and destablizing, increase. An update policy that does result in a linear increase of ....

PRUE,W.,AND POSTEL,J. Something A Host Could Do with Source Quench. ARPANET Working Group Requests for Comment, DDNNetwork Information Center, SRI International, Menlo Park, CA, July 1987. RFC-1016.

....to gateway mechanisms for congestion avoidance Early descriptions of IP Source Quench messages suggest that gateways could send Source Quench messages to source hosts before the buffer space at the gateway reaches capacity [26] and before packets have to be dropped at the gateway. One proposal [27] suggests that the gateway send Source Quench messages when the queue size exceeds a certain threshold, and outlines a possible method for flow control at the source hosts in response to these messages. The proposal also suggests that when the gateway queue size approaches the maximum level the ....

Prue, W., and Postel, J., "Something a Host Could Do with Source Quench", RFC 1016, July 1987.

....the best convergence properties. For this study, as a representative of binary feedback mechanisms with a AIMD window variation policy we chose a mechanism introduced by Acharya et. al (ERAF [29, 22] A similar mechanism, but using a additive increase additivedecrease window discipline (SquId [28]) was proposed by Prue Postel. Recent work by Floyd [15] examined alternate window variation policies, namely TCP friendly rate control mechanisms. We do not speci cally consider TCP friendly 5 window variation policies and for a comparison of TRFC and AIMD the reader is referred to [16] 4.1 ....

W. Prue and J. Postel. Something a host could do with a source quench. RFC 1016, Jul 1987.

....receipt of a single ECN message serves as a notification of congestion to the TCP source. At the same time, the guidelines ensure that the TCP source does not respond to ECN messages more frequently than necessary. Guidelines: ffl TCP s response to ECN should be similar, over 1 As an example, [PP87] suggested that the gateway send Source Quench messages when the queue size exceeds a certain threshold. longer time scales, to its response to a dropped packet as an indication of congestion. ffl Over smaller time scales (e.g. one or two round trip times) TCP s response to ECN can be less ....

Prue, W., and Postel, J., "Something a Host Could Do with Source Quench", Request for Comments (RFC) 1016, July 1987.

....and a source quench generated for any flow in an aggregate results in a decrease in the regulator bandwidth for the entire aggregate. We evaluate the impact of using aggregation in Section 4.4. The ERUF algorithm is similar, in spirit, to the SQuID algorithm proposed by Prue Postel in RFC 1016 [25]. SQuID specifies end host response to source quenches. There are, however, several differences. First, the SQuID algorithm adjusts the per packet regulator delay instead of the regulator bandwidth. For flows with variable packet size, this can lead to undesirable variations in the sending rate. ....

W. Prue and J. Postel. Something a host could do with a source quench. RFC 1016, Jul 1987.

....to gateway mechanisms for congestion avoidance Early descriptions of IP Source Quench messages suggest that gateways could send Source Quench messages to source hosts before the buffer space at the gateway reaches capacity [26] and before packets have to be dropped at the gateway. One proposal [27] suggests that the gateway send Source Quench messages when the queue size exceeds a certain threshold, and outlines a possible method for flow control at the source hosts in response to these messages. The proposal also suggests that when the gateway queue size approaches the maximum level the ....

Prue, W., and Postel, J., "Something a Host Could Do with Source Quench", RFC 1016, July 1987.

....receipt of a single ECN message serves as a notification of congestion to the TCP source. At the same time, the guidelines ensure that the TCP source does not respond to ECN messages more frequently than necessary. Guidelines: ffl TCP s response to ECN should be similar, over 1 As an example, [PP87] suggested that the gateway send Source Quench messages when the queue size exceeds a certain threshold. longer time scales, to its response to a dropped packet as an indication of congestion. ffl Over smaller time scales (e.g. one or two round trip times) TCP s response to ECN can be less ....

Prue, W., and Postel, J., "Something a Host Could Do with Source Quench", Request for Comments (RFC) 1016, July 1987.

....an aggregate results in a decrease in the regulator bandwidth for the entire aggregate. We evaluate the impact of using aggregation in Section 3. The ERUF algorithm is similar, in spirit, to the SQuID algorithm proposed by Prue Postel in RFC 1016 for specifying end host response to source quenches [24]. There are, however, several differences. First, the SQuID algorithm adjusts the per packet regulator delay instead of the regulator bandwidth. For flows with variable packet size, this can lead to undesirable variations in the sending rate. Second, the SQuID algorithm uses a ....

W. Prue and J. Postel. Something a host could do with a source quench. RFC 1016, Jul 1987.

....the algorithm follows immediately: The multiplicative decrease remains a multiplicative decrease (e.g. double the interval between packets) But subtracting a constant amount from interval does not result in an additive increase in bandwidth. This approach has been tried, e.g. 18] and [24], and appears to oscillate badly. To see why, note that for an inter packet interval I and decrement c , the bandwidth change of a decrease interval by constant policy is 1 I 1 I 0 c a non linear, and destablizing, increase. An update policy that does result in a linear increase of ....

Prue, W., and Postel, J. Something A Host Could Do with Source Quench. Arpanet Working Group Requests for Comment, DDN Network Information Center, SRI International, Menlo Park, CA, July 1987. RFC-1016.

....to gateway mechanisms for congestion avoidance Early descriptions of IP Source Quench messages suggest that gateways could send Source Quench messages to source hosts before the buffer space at the gateway reaches capacity [26] and before packets have to be dropped at the gateway. One proposal [27] suggests that the gateway send Source Quench messages when the queue size exceeds a certain threshold, and outlines a possible method for flow control at the source hosts in response to these messages. The proposal also suggests that when the gateway queue size approaches the maximum level the ....

Prue, W., and Postel, J., "Something a Host Could Do with Source Quench", RFC 1016, July 1987.

No context found.

Prue, W., and Postel, J., "Something a Host Could Do with Source Quench", Request for Comments (RFC) 1016, July 1987.

No context found.

PRUE, W., AND POSTEL, J. Something A Host Could Do with Source Quench. ARPANET Working Group Requests for Comment, DDN Network Information Center, SRI International, Menlo Park, CA, July 1987. RFC-1016.

No context found.

Prue, W., and Postel, J., "Something a Host Could Do with Source Quench", RFC 1016, July 1987.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC