< draft-ietf-pwe3-congcons-02.txt | draft-ietf-pwe3-congcons-03a.txt > | |||
---|---|---|---|---|
PWE3 YJ. Stein | PWE3 YJ. Stein | |||
Internet-Draft RAD Data Communications | Internet-Draft RAD Data Communications | |||
Intended status: Informational D. Black | Intended status: Informational D. Black | |||
Expires: January 25, 2015 EMC Corporation | Expires: February 15, 2015 EMC Corporation | |||
B. Briscoe | B. Briscoe | |||
BT | BT | |||
July 24, 2014 | August 14, 2014 | |||
Pseudowire Congestion Considerations | Pseudowire Congestion Considerations | |||
draft-ietf-pwe3-congcons-02 | draft-ietf-pwe3-congcons-03 | |||
Abstract | Abstract | |||
Pseudowires (PWs) have become a common mechanism for tunneling | Pseudowires (PWs) have become a common mechanism for tunneling | |||
traffic, and may be found in unmanaged scenarios competing for | traffic, and may be found in unmanaged scenarios competing for | |||
network resources both with other PWs and with non-PW traffic, such | network resources both with other PWs and with non-PW traffic, such | |||
as TCP/IP flows. It is thus worthwhile specifying under what | as TCP/IP flows. It is thus worthwhile specifying under what | |||
conditions such competition is safe, i.e., the PW traffic does not | conditions such competition is safe, i.e., the PW traffic does not | |||
significantly harm other traffic or contribute more than it should to | significantly harm other traffic or contribute more than it should to | |||
congestion. We conclude that PWs transporting responsive traffic | congestion. We conclude that PWs transporting responsive traffic | |||
behave as desired without the need for additional mechanisms. For | behave as desired without the need for additional mechanisms. For | |||
inelastic PWs (such as TDM PWs) we derive a bound under which such | inelastic PWs (such as TDM PWs) we derive a bound under which such | |||
PWs consume no more network capacity than a TCP flow. We also | PWs consume no more network capacity than a TCP flow. We also | |||
propose employing a transport circuit breaker | propose employing a transport circuit breaker that shuts down a TDM | |||
[I-D.fairhurst-tsvwg-circuit-breaker] that shuts down a TDM PW | PW consistently surpassing this bound, as the emulated TDM service | |||
consistently surpassing this bound, as the emulated TDM service | ||||
itself would be be of insufficient quality. | itself would be be of insufficient quality. | |||
Status of This Memo | Status of This Memo | |||
This Internet-Draft is submitted in full conformance with the | This Internet-Draft is submitted in full conformance with the | |||
provisions of BCP 78 and BCP 79. | provisions of BCP 78 and BCP 79. | |||
Internet-Drafts are working documents of the Internet Engineering | Internet-Drafts are working documents of the Internet Engineering | |||
Task Force (IETF). Note that other groups may also distribute | Task Force (IETF). Note that other groups may also distribute | |||
working documents as Internet-Drafts. The list of current Internet- | working documents as Internet-Drafts. The list of current Internet- | |||
Drafts is at http://datatracker.ietf.org/drafts/current/. | Drafts is at http://datatracker.ietf.org/drafts/current/. | |||
Internet-Drafts are draft documents valid for a maximum of six months | Internet-Drafts are draft documents valid for a maximum of six months | |||
and may be updated, replaced, or obsoleted by other documents at any | and may be updated, replaced, or obsoleted by other documents at any | |||
time. It is inappropriate to use Internet-Drafts as reference | time. It is inappropriate to use Internet-Drafts as reference | |||
material or to cite them other than as "work in progress." | material or to cite them other than as "work in progress." | |||
This Internet-Draft will expire on January 25, 2015. | This Internet-Draft will expire on February 15, 2015. | |||
Copyright Notice | Copyright Notice | |||
Copyright (c) 2014 IETF Trust and the persons identified as the | Copyright (c) 2014 IETF Trust and the persons identified as the | |||
document authors. All rights reserved. | document authors. All rights reserved. | |||
This document is subject to BCP 78 and the IETF Trust's Legal | This document is subject to BCP 78 and the IETF Trust's Legal | |||
Provisions Relating to IETF Documents | Provisions Relating to IETF Documents | |||
(http://trustee.ietf.org/license-info) in effect on the date of | (http://trustee.ietf.org/license-info) in effect on the date of | |||
publication of this document. Please review these documents | publication of this document. Please review these documents | |||
skipping to change at page 2, line 40 | skipping to change at page 2, line 40 | |||
Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 22 | Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 22 | |||
1. Introduction | 1. Introduction | |||
A pseudowire (PW)(see [RFC3985]) is a construct for tunneling a | A pseudowire (PW)(see [RFC3985]) is a construct for tunneling a | |||
native service, such as Ethernet or TDM, over a Packet Switched | native service, such as Ethernet or TDM, over a Packet Switched | |||
Network (PSN), such as IPv4, IPv6, or MPLS. The PW packet | Network (PSN), such as IPv4, IPv6, or MPLS. The PW packet | |||
encapsulates a unit of native service information by prepending the | encapsulates a unit of native service information by prepending the | |||
headers required for transport in the particular PSN (which must | headers required for transport in the particular PSN (which must | |||
include a demultiplexer field to distinguish the different PWs) and | include a demultiplexer field to distinguish the different PWs) and | |||
preferably the 4 byte PWE3 control word. | preferably the 4 byte Pseudowire Emulation Edge-to-Edge (PWE3) | |||
control word. | ||||
PWs have no bandwidth reservation or control mechanisms, meaning that | PWs have no bandwidth reservation or control mechanisms, meaning that | |||
when multiple PWs are transported in parallel, and/or in parallel | when multiple PWs are transported in parallel, and/or in parallel | |||
with other flows, there is no defined means for allocating resources | with other flows, there is no defined means for allocating resources | |||
for any particular PW, or for preventing negative impact of a | for any particular PW, or for preventing negative impact of a | |||
particular PW on neighboring flows. Mechanisms for provisioning PWs | particular PW on neighboring flows. The case where the service | |||
in service provider networks are well understood and will not be | provider network arranges sufficient capacity for a PW is well | |||
discussed further here. | understood and will not be discussed further here. The concern is | |||
over PWs deployed independently of the service provider network's | ||||
traffic engineering or capacity planning. | ||||
While PWs are most often placed in MPLS tunnels, there are several | While PWs are most often placed in MPLS tunnels, there are several | |||
mechanisms that enable transporting PWs over an IP infrastructure. | mechanisms that enable PWs to be transported over an IP | |||
These include: | infrastructure. These include: | |||
UDP/IP encapsulations defined for TDM PWs | o UDP/IP encapsulations defined for PWs emulating time division | |||
([RFC4553][RFC5086][RFC5087]), | multiplexing (TDM) ([RFC4553][RFC5086][RFC5087]), | |||
L2TPv3 based PWs, | o L2 tunnelling protocol (L2TPv3) based PWs, | |||
MPLS PWs directly over IP according to RFC 4023 [RFC4023], | o MPLS PWs directly over IP according to RFC 4023 [RFC4023], | |||
MPLS PWs over GRE over IP according to RFC 4023 [RFC4023]. | o MPLS PWs over Generic Routing Encapsulation (GRE) over IP | |||
according to RFC 4023 [RFC4023]. | ||||
Whenever PWs are transported over IP, they may compete for network | Whenever PWs are transported over IP, they may compete for network | |||
resources with neighboring congestion-responsive flows (e.g., TCP | resources with neighboring congestion-responsive flows (e.g., TCP | |||
flows). In this document we study the effect of PWs on such | flows). In this document we study the effect of PWs on such | |||
neighboring flows, and discover that the negative impact of PW | neighboring flows, and discover that the negative impact of PW | |||
traffic is generally no worse than that of congestion-responsive | traffic is generally no worse than that of congestion-responsive | |||
flows, ([RFC2914],[RFC5033]}. | flows ([RFC2914],[RFC5033]}. | |||
At first glance one may consider a PW transported over IP to be | At first glance one may consider a PW transported over IP to be | |||
considered as a single flow, on a par with a single TCP flow. Were | considered as a single flow, on a par with a single TCP flow. Were | |||
we to accept this tenet, we would require a PW to back off under | we to accept this tenet, we would require a PW to back off under | |||
congestion to consume no more bandwidth than a single TCP flow under | congestion to consume no more bandwidth than a single TCP flow under | |||
such conditions (see [RFC5348]). However, since PWs may carry | such conditions (see [RFC5348]). However, since PWs may carry | |||
traffic from many users, it makes more sense to consider each PW to | traffic from many users, it makes more sense to consider each PW to | |||
be equivalent to multiple TCP flows. | be equivalent to multiple TCP flows. | |||
The following two sections consider PWs of two types. | The following two sections consider PWs of two types. | |||
Elastic Flows: Section 2 concludes that the response to congestion | Elastic Flows: Section 2 concludes that the response to congestion | |||
of a PW carrying elastic (e.g., TCP) flows is no different to the | of a PW carrying elastic (e.g., TCP) flows is no different to the | |||
combined behaviour of the set of the same elastic flows were they | combined individual behaviours of the set of the same elastic | |||
not encapsulated within a PW. | flows were they not encapsulated within a PW. | |||
Inelastic Flows: Section 3 considers the case of inelastic constant | Inelastic Flows: Section 3 considers the case of inelastic constant | |||
bit-rate (CBR) TDM PWs ([RFC4553][RFC5086] [RFC5087]) competing | bit-rate (CBR) TDM PWs ([RFC4553][RFC5086] [RFC5087]) competing | |||
with TCP flows. Such PWs require a preset amount of bandwidth, | with TCP flows. Such PWs require a preset amount of bandwidth, | |||
that may be lower or higher than that consumed by an otherwise | that may be lower or higher than that consumed by an otherwise | |||
unconstrained TCP flow under the same network conditions. In any | unconstrained TCP flow under the same network conditions. In any | |||
case, such a PW is inable to respond to congestion in a TCP-like | case, such a PW is unable to respond to congestion in a TCP-like | |||
manner; on the other hand, the total bandwidth it consumes remains | manner; although at least the total bandwidth it consumes remains | |||
constant and does not increase to consume additional bandwidth as | constant and does not increase to consume additional bandwidth as | |||
TCP rates back off. If the bandwidth consumed by a TDM PW is | TCP rates back off. If the bandwidth consumed by a TDM PW is | |||
considered detrimental, the only available remedy is to completely | considered detrimental, the only available remedy is to completely | |||
shut down the PW, by using a transport circuit breaker mechanism. | shut down the PW, by using a transport circuit breaker mechanism | |||
However, we will show that even before such an action is | [I-D.fairhurst-tsvwg-circuit-breaker]. However, we will show that | |||
warranted, the PW will become unable to faithfully emulate the | even before such an action is warranted, the PW will become unable | |||
native TDM service; for example, when a TDM service is carrying | to faithfully emulate the native TDM service; for example, when a | |||
voice grade telephony channels, the voice quality will degrade to | TDM service is carrying voice grade telephony channels, the voice | |||
below useful levels. | quality will degrade to below useful levels. | |||
Thus, in both cases, pseudowires will not inflict significant harm on | Thus, in both cases, pseudowires will not inflict significant harm on | |||
neighboring TCP flows, as in one case they respond adequately to | neighboring TCP flows, as in one case they respond adequately to | |||
congestion, and in the other they would be shut down due to being | congestion, and in the other they would be shut down due to being | |||
unable to emulate the native service before harming neighboring | unable to emulate the native service before harming neighboring | |||
flows. | flows. | |||
2. PWs Comprising Elastic Flows | 2. PWs Comprising Elastic Flows | |||
In this section we consider Ethernet PWs that primarily carry | In this section we consider Ethernet PWs that primarily carry | |||
skipping to change at page 5, line 9 | skipping to change at page 5, line 9 | |||
constituent TCP flows. In addition, the individual TCP flows would | constituent TCP flows. In addition, the individual TCP flows would | |||
still back off due to their end points being oblivious to the fact | still back off due to their end points being oblivious to the fact | |||
that they are carried over a PW. This would further degrade the | that they are carried over a PW. This would further degrade the | |||
flow's throughput as compared to a non-PW-encapsulated flow, in | flow's throughput as compared to a non-PW-encapsulated flow, in | |||
contradiction to desirable behavior. | contradiction to desirable behavior. | |||
3. PWs Comprising Inelastic Flows | 3. PWs Comprising Inelastic Flows | |||
Inelastic PWs, such as TDM PWs ([RFC4553][RFC5086][RFC5087]), are | Inelastic PWs, such as TDM PWs ([RFC4553][RFC5086][RFC5087]), are | |||
potentially more problematic than the elastic PWs of the previous | potentially more problematic than the elastic PWs of the previous | |||
section. Being constant bit-rate (CBR), TDM PWs can not be made | section. Being constant bit-rate (CBR), TDM PWs are not responsive | |||
responsive to congestion. On the other hand, being CBR, they also do | to congestion. On the other hand, being CBR, they at least do not | |||
not attempt to capture additional bandwidth when neighboring TCP | attempt to capture additional bandwidth when neighboring TCP flows | |||
flows back off. | back off. | |||
Since a TDM PW continuously consumes a constant amount of bandwidth, | Since a TDM PW continuously consumes a constant amount of bandwidth, | |||
if the bandwidth occupied by a TDM PW endangers the network as a | if the bandwidth occupied by a TDM PW endangers the network as a | |||
whole, the only recourse is to shut it down, denying service to all | whole, the only recourse is to shut it down, denying service to all | |||
customers of the TDM native service. We can accomplish this by | customers of the TDM native service. We can accomplish this by | |||
employing a transport circuit breaker, by which we mean an automatic | employing a transport circuit breaker, by which we mean an automatic | |||
mechanism for terminating a flow to prevent negative impact on other | mechanism for terminating a flow to prevent negative impact on other | |||
flows and on the stability of the network | flows and on the performance of the network | |||
[I-D.fairhurst-tsvwg-circuit-breaker]. Note that a transport circuit | [I-D.fairhurst-tsvwg-circuit-breaker]. Note that a transport circuit | |||
breaker is intended as a protection mechanism of last resort, just as | breaker is intended as a protection mechanism of last resort, just as | |||
an electrical circuit breaker is only triggered when absolutely | an electrical circuit breaker is only triggered when absolutely | |||
necessary. We should mention in passing that under certain | necessary. We should mention in passing that under certain | |||
conditions it may be possible to reduce the bandwidth consumption of | conditions it may be possible to reduce the bandwidth consumption of | |||
a TDM PW. A prevalent case is that of a TDM native service that | a TDM PW. A prevalent case is that of a TDM native service that | |||
carries voice channels that may not all be active. Using the AAL2 | carries voice channels that may not all be active. Using the AAL2 | |||
mode of [RFC5087] (perhaps along with connection admission control) | mode of [RFC5087] (perhaps along with connection admission control) | |||
can enable bandwidth adaptation, at the expense of more sophisticated | can enable bandwidth adaptation, at the expense of more sophisticated | |||
native service processing (NSP). | native service processing (NSP). | |||
In the following we will show that for many cases of interest a TDM | In the following we will show that for many cases of interest a TDM | |||
PW, treated as a single flow, will behave in a reasonable manner | PW, even treated as a single flow, will behave in a reasonable manner | |||
without any additional mechanisms. We will focus on structure- | without any additional mechanisms. We will focus on structure- | |||
agnostic TDM PWs [RFC4553] although our analysis can be readily | agnostic TDM PWs [RFC4553] although our analysis can be readily | |||
applied to structure-aware PWs (see Appendix A). | applied to structure-aware PWs (see Appendix A). | |||
In order to quantitatively compare TDM PWs to TCP flows, we will | In order to quantitatively compare TDM PWs to TCP flows, we will | |||
compare the effect of TDM PW packets with that of TCP packets of the | compare the effect of TDM PW traffic with that of TCP traffic of the | |||
same packet size and sent at the same rate. This is potentially an | same packet size and delay. For PWs this is potentially an overly | |||
overly pessimistic comparison, as TDM PW packets are frequently | pessimistic comparison, as TDM PW packets are frequently configured | |||
configured to be short in order to minimize latency, while TCP | to be short in order to minimize latency, while TCP packets are free | |||
packets are free to be much larger. | to be much larger. | |||
There are two network parameters relevant to our discussion, namely | There are two network parameters relevant to our discussion, namely | |||
the one-way delay D and the packet loss rate PLR. The one-way delay | the one-way delay (D) and the packet loss probability (PLP). The | |||
of a native TDM service consists of the physical time-of-flight plus | one-way delay of a native TDM service consists of the physical time- | |||
125 microseconds for each TDM switch traversed; and is thus very | of-flight plus 125 microseconds for each TDM switch traversed; so the | |||
small as compared to typical PSN network-crossing latencies. Many | total one-way delay for most TDM PWs has to be small compared to | |||
protocols and applications running over TDM circuits thus expect | typical PSN network-crossing latencies. Many protocols and | |||
extremely low delay, and thus in our comparisons we will only | applications running over TDM circuits thus expect extremely low | |||
consider delays of a few milliseconds. | delay, and thus in our comparisons we will only consider delays of a | |||
few milliseconds. | ||||
Regarding packet loss, the TDM PW RFCs specify behaviors upon | Regarding packet loss, the TDM PW RFCs specify the appropriate | |||
detecting a lost packet. Structure-agnostic transport has no | behavior upon detecting a lost packet: | |||
alternative to outputting an "all-ones" AIS pattern towards the TDM | ||||
circuit, which, when long enough in duration, is recognized by the | Structure-agnostic transport has no alternative to outputting an | |||
receiving TDM device as a fault indication (see Appendix A). | "all-ones" AIS pattern towards the TDM circuit, which, when long | |||
International standards place stringent limits on the number of such | enough in duration, is recognized by the receiving TDM device as a | |||
faults tolerated. Calculations presented in the appendix show that | fault indication (see Appendix A). ITU standard [G826] places | |||
only loss probabilities in the realm of fractions of a percent are | stringent limits on the number of such faults tolerated. | |||
relevant for structure-agnostic transport (see Appendix A). | Calculations presented in the appendix show that only loss | |||
probabilities in the realm of fractions of a percent are relevant | ||||
for structure-agnostic transport (see Appendix A). | ||||
Structure-aware transport regenerates frame alignment signals thus | Structure-aware transport regenerates frame alignment signals thus | |||
hiding AIS indications resulting from infrequent packet loss. | hiding AIS indications resulting from infrequent packet loss. | |||
Furthermore, for TDM circuits carrying voice channels the use of | Furthermore, for TDM circuits carrying voice channels the use of | |||
packet loss concealment algorithms is possible (such algorithms have | packet loss concealment algorithms is possible (such algorithms | |||
been previously described for TDM PWs). However, even structure- | have been previously described for TDM PWs). However, even | |||
aware transport ceases to provide a useful service at about 2 percent | structure-aware transport ceases to provide a useful service at | |||
loss probability. Hence, in our comparisons we will only consider | about 2 percent loss probability (see Appendix A). Hence, in our | |||
PLRs of 1 or 2 percent. | comparisons we will only consider PLPs of 1 or 2 percent. | |||
RFC 5348 on TCP Friendly Rate Control (TFRC) [RFC5348] provides a | RFC 5348 on TCP Friendly Rate Control (TFRC) [RFC5348] provides a | |||
simplified formula for TCP throughput as a function of delay and | simplified formula for TCP throughput as a function of delay and | |||
packet loss rate. | packet loss probability. | |||
S | S | |||
X = ------------------------------------------------ | X = ------------------------------------------------ | |||
R ( sqrt(2p/3) + 12 sqrt(3p/8) p (1+32p^2) ) | R * ( sqrt(2*p/3) + 12*sqrt(3*p/8)*p*(1+32*p^2) ) | |||
where | where | |||
X is average sending rate in Bytes per second, | X is average sending rate in Bytes per second, | |||
S is the segment (packet payload) size in Bytes, | S is the segment (packet payload) size in Bytes, | |||
R is the round-trip time in seconds, | R is the round-trip time in seconds, | |||
p is the packet loss probability (i.e., PLR/100). | p is the packet loss probability (i.e., PLP/100). | |||
We can now compare the bandwidth consumed by TDM pseudowires with | We can now compare the bandwidth consumed by TDM pseudowires with | |||
that of a TCP flow for given packet loss and delay. The results are | that of a TCP flow for given packet loss and delay. The results are | |||
depicted in the accompanying figures (available only in the PDF | depicted in the accompanying figures (available only in the PDF | |||
version of this document). In Figures 1 and 2 we see the | version of this document). In Figures 1 and 2 we see the | |||
conventional rate vs. packet loss plot for low-rate TDM (both T1 and | conventional rate vs. packet loss plot for low-rate TDM (both T1 and | |||
E1) traffic, as well as TCP traffic with the same payload size (64 or | E1) traffic, as well as TCP traffic with the same payload size (64 or | |||
256 Bytes respectively). Since the TDM rates are constant (T1 and E1 | 256 Bytes respectively). Since the TDM rates are constant (T1 and E1 | |||
having payload throughputs of 1.544 Mbps and 2.048 Mbps | having payload throughputs of 1.544 Mbps and 2.048 Mbps | |||
respectively), and the TDM service can only be faithfully emulated | respectively), and the TDM service can only be faithfully emulated | |||
using SAToP up to a PLR of about half a percent, the T1 and E1 | using Structure-Agnostic TDM over packet (SAToP) up to a PLP of about | |||
pseudowires occupy line segments on the graph. On the other hand, | half a percent, the T1 and E1 pseudowires occupy line segments on the | |||
the TCP rate equation produces rate curves dependent on both delay | graph. On the other hand, the TCP rate equation produces rate curves | |||
and packet loss. | dependent on both delay and packet loss. | |||
We see that in general for large packet sizes, short delays, and low | We see that in general for large packet sizes, short delays, and low | |||
packet loss rates, the TDM pseudowires consume much less bandwidth | packet loss probabilities, the TDM pseudowires consume much less | |||
than TCP would under identical conditions. Only for small packets, | bandwidth than TCP would under identical conditions. Only for small | |||
long delays, and high packet loss ratios, do TDM PWs potentially | packets, long delays, and high packet loss ratios, do TDM PWs | |||
consume more bandwidth, and even then only marginally. Similarly, in | potentially consume more bandwidth, and even then only marginally. | |||
Figures 3 and 4 we repeat the exercise for higher rate E3 and T3 | Further, recall that the use of small packets by TCP traffic is only | |||
(rates 34.368 and 44.736 Mbps respectively) pseudowires, allowing | for this worst-case analysis, and in practice TCP does not need to to | |||
delays and PLRs suitable appropriate for these signals. We see that | use the same small packets that competing PW traffic might need to | |||
the TDM pseudowires consume much less bandwidth than TCP, for all | use. Similarly, in Figures 3 and 4 we repeat the exercise for higher | |||
reasonable parameter combinations. | rate E3 and T3 (rates 34.368 and 44.736 Mbps respectively) | |||
pseudowires, allowing delays and PLPs suitable for these signals (see | ||||
Appendix A). We see that the TDM pseudowires consume much less | ||||
bandwidth than TCP, for all reasonable parameter combinations. | ||||
-------------------------------------------------------------------- | -------------------------------------------------------------------- | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I E1/T1 PWs vs. TCP for segment size 64B I | I E1/T1 PWs vs. TCP for segment size 64B I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
skipping to change at page 11, line 23 | skipping to change at page 11, line 23 | |||
D < ----------- | D < ----------- | |||
BW f(p) | BW f(p) | |||
where | where | |||
D is the one-way delay, | D is the one-way delay, | |||
S is the TDM segment size (packet excluding overhead) in Bytes, | S is the TDM segment size (packet excluding overhead) in Bytes, | |||
BW is TDM service bandwidth in bits per second, | BW is TDM service bandwidth in bits per second, | |||
f(p) = sqrt(2p/3) + 12 sqrt(3p/8) p (1+32p^2). | f(p) = sqrt(2p/3) + 12 sqrt(3p/8) p (1+32p^2). | |||
One may view this condition as defining an operating envelope for a | One may view this condition as defining an undeniably safe operating | |||
TDM PW, as a TDM PW that occupies no more bandwidth than a TCP flow | envelope for a TDM PW, as a TDM PW that occupies no more bandwidth | |||
causes no more congestion than that TCP flow would. Under this | than a TCP flow causes no more congestion than that TCP flow would. | |||
condition it is safe to place the TDM PW along with congestion- | This does not draw us into a debate over whether TCP-friendliness is | |||
responsive traffic such as TCP, without causing additional | a valid constraint; it simply says that there can be no question that | |||
congestion. on the other hand, were the TDM PW to consume | a PW is safe if it does no more harm than a single TCP flow. Under | |||
significantly more bandwidth a TCP flow, it could contribute | this condition it is undeniably safe to place the TDM PW along with | |||
disproportionately to congestion, and its mixture with congestion- | congestion- responsive traffic such as TCP. On the other hand, were | |||
responsive traffic might be inappropriate. | the TDM PW to consume significantly more bandwidth a TCP flow, it | |||
could contribute disproportionately to congestion, and its mixture | ||||
with congestion- responsive traffic might start to cause concern. | ||||
We see in Figures 5 and 6 that TDM pseudowires carrying T1 or E1 | ||||
native services satisfy the condition for all parameters of interest | ||||
for large packet sizes (e.g., S=512 Bytes of TDM data). For the | ||||
SAToP default of 256 Bytes, as long as the one-way delay is less than | ||||
10 milliseconds, the loss probability can exceed respectively 0.6 or | ||||
0.3 percent. For packets containing 128 or 64 Bytes the constraints | ||||
are more troublesome, but there are still parameter ranges where the | ||||
TDM PW consumes less than a TCP flow under similar conditions. | ||||
Similarly, Figures 7 and 8 demonstrate that E3 and T3 native services | ||||
with the SAToP default of 1024 Bytes of TDM per packet satisfy the | ||||
condition for a broad spectrum of delays and PLPs. | ||||
We derived this condition assuming steady-state conditions, and thus | We derived this condition assuming steady-state conditions, and thus | |||
two caveats are in order. First, the condition does not specify how | two caveats are in order. First, the condition does not specify how | |||
to treat a TDM PW that initially satisfies the condition, but is then | to treat a TDM PW that initially satisfies the condition, but is then | |||
faced with a deteriorating network environment. In such cases one | faced with a deteriorating network environment. In such cases one | |||
additionally needs to analyze the reaction times of the responsive | additionally needs to analyze the reaction times of the responsive | |||
flows to congestion events. Second, the derivation assumed that the | flows to congestion events. Second, the derivation assumed that the | |||
TDM PW was competing with long-lived TDM flows, because under this | TDM PW was competing with long-lived TCP flows, because under this | |||
assumption it was straightforward to obtain a quantitative comparison | assumption it was straightforward to obtain a quantitative comparison | |||
with something widely considered to offer a safe response to | with something widely considered to offer a safe response to | |||
congestion. Short-lived TCP flows may find themselves disadvantaged | congestion. Short-lived TCP flows may find themselves disadvantaged | |||
as compared to a long-lived TDM PW satisfying the condition. | as compared to a long-lived TDM PW satisfying the condition. | |||
We see in Figures 5 and 6 that TDM pseudowires carrying T1 or E1 | ||||
native services satisfy the condition for all parameters of interest | ||||
for large packet sizes (e.g., S=512 Bytes of TDM data). For the | ||||
SAToP default of 256 Bytes, as long as the one-way delay is less than | ||||
10 milliseconds, the loss probability can exceed 0.3 or 0.6 percent. | ||||
For packets containing 128 or 64 Bytes the constraints are more | ||||
troublesome, but there are still parameter ranges where the TDM PW | ||||
consumes less than a TCP flow under similar conditions. Similarly, | ||||
Figures 7 and 8 demonstrate that E3 and T3 native services with the | ||||
SAToP default of 1024 Bytes of TDM per packet satisfy the condition | ||||
for a broad spectrum of delays and PLRs. | ||||
Note that violating the condition for a short amount of time is not | Note that violating the condition for a short amount of time is not | |||
sufficient justification for shutting down the TDM PW. While TCP | sufficient justification for shutting down the TDM PW. While TCP | |||
flows react within a round trip time, PW commissioning and | flows react within a round trip time, PW commissioning and | |||
decommissioning are time consuming processes that should only be | decommissioning are time consuming processes that should only be | |||
undertaken when it becomes clear that the congestion is not | undertaken when it becomes clear that the congestion is not | |||
transient. | transient. | |||
-------------------------------------------------------------------- | -------------------------------------------------------------------- | |||
I I | I I | |||
I I | I I | |||
skipping to change at page 18, line 23 | skipping to change at page 18, line 23 | |||
Appendix A. Loss Probabilities for TDM PWs | Appendix A. Loss Probabilities for TDM PWs | |||
ITU-T Recommendation G.826 [G826] specifies limits on the Errored | ITU-T Recommendation G.826 [G826] specifies limits on the Errored | |||
Second Ratio (ESR) and the Severely Errored Second Ratio (SESR). For | Second Ratio (ESR) and the Severely Errored Second Ratio (SESR). For | |||
our purposes, we will simplify the definitions and understand an | our purposes, we will simplify the definitions and understand an | |||
Errored Second (ES) to be a second of time during which a TDM bit | Errored Second (ES) to be a second of time during which a TDM bit | |||
error occurred or a defect indication was detected. A Severely | error occurred or a defect indication was detected. A Severely | |||
Errored Second (SES) is an ES second during which the Bit Error Rate | Errored Second (SES) is an ES second during which the Bit Error Rate | |||
(BER) exceeded one in one thousand (10^-3). Note that if the error | (BER) exceeded one in one thousand (10^-3). Note that if the error | |||
condition AIS was detected according to the criteria of ITU-T | condition Alarm Indication Signal (AIS) was detected according to the | |||
Recommendation G.775 [G826] a SES was considered to have occurred. | criteria of ITU-T Recommendation G.775 [G775] a SES was considered to | |||
The respective ratios are the fraction of ES or SES to the total | have occurred. The respective ratios are the fraction of ES or SES | |||
number of seconds in the measurement interval. | to the total number of seconds in the measurement interval. | |||
For both E1 and T1 TDM circuits, G.826 allows ESR of 4% (0.04), and | For both E1 and T1 TDM circuits, G.826 allows ESR of 4% (0.04), and | |||
SESR of 1/5% (0.002). For E3 and T3 the ESR must be no more than | SESR of 0.2% (0.002). For E3 and T3 the ESR must be no more than | |||
7.5% (0.075), while the SESR is unchanged. | 7.5% (0.075), while the SESR is unchanged. | |||
Focusing on E1 circuits, the ESR of 4% translates, assuming the worst | Focusing on E1 circuits, the ESR of 4% translates, assuming the worst | |||
case of isolated exactly periodic packet loss, to a packet loss event | case of isolated exactly periodic packet loss, to a packet loss event | |||
no more than every 25 seconds. However, once a packet is lost, | no more than every 25 seconds. However, once a packet is lost, | |||
another packet lost in the same second doesn't change the ESR, | another packet lost in the same second doesn't change the ESR, | |||
although it may contribute to the ES becoming a SES. Assuming an | although it may contribute to the ES becoming a SES. E1 circuits run | |||
integer number of TDM frames per PW packet, the number of packets per | at 8000 frames per second. Therefore, assuming an integer number of | |||
second is given by packets per second = 8000 / (frames per packet), | TDM frames per PW packet, the number of packets per second is given | |||
where prevalent cases are 1, 2, 4 and 8 frames per packet. Since for | by packets per second = 8000 / (frames per packet), where prevalent | |||
these cases there will be 8000, 4000, 2000, and 1000 packets per | cases are 1, 2, 4 and 8 frames per packet. Since for these cases | |||
second, respectively, the maximum allowed packet loss probability is | there will be 8000, 4000, 2000, and 1000 packets per second, | |||
0.0005%, 0.001%, 0.002%, and 0.004% respectively. | respectively, the maximum allowed packet loss probability is 0.0005%, | |||
0.001%, 0.002%, and 0.004% respectively. | ||||
These extremely low allowed packet loss probabilities are only for | These extremely low allowed packet loss probabilities are only for | |||
the worst case scenario. In reality, when packet loss is above | the worst case scenario. With tail-drop buffers, when packet loss is | |||
0.001%, it is likely that loss bursts will occur. If the lost | above 0.001%, it is likely that loss bursts will occur. If the lost | |||
packets are sufficiently close together (we ignore the precise | packets are sufficiently close together (we ignore the precise | |||
details here) then the permitted packet loss rate increases by the | details here) then the permitted packet loss probability increases by | |||
appropriate factor, without G.826 being cognizant of any change. | the appropriate factor, without G.826 being cognizant of any change. | |||
Hence the worst-case analysis is expected to be extremely pessimistic | Hence the worst-case analysis is expected to be extremely pessimistic | |||
for real networks. Next we will go to the opposite extreme and | for networks with tail-drop buffers, although it will be closer to | |||
assume that all packet loss events are in periodic loss bursts. In | reality in networks with active queue management widely deployed. | |||
order to minimize the ESR we will assume that the burst lasts no more | Next we will go to the opposite extreme and assume that all packet | |||
than one second, and so we can afford to lose no more than packet per | loss events are in periodic loss bursts. In order to minimize the | |||
second packets in each burst. As long as such one-second bursts do | ESR we will assume that the burst lasts no more than one second, and | |||
not exceed four percent of the time, we still maintain the allowable | so we can afford to lose no more than {something missing here?} | |||
ESR. Hence the maximum permissible packet loss rate is 4%. Of | packet per second packets in each burst. As long as such one-second | |||
course, this estimate is extremely optimistic, and furthermore does | bursts do not exceed four percent of the time, we still maintain the | |||
not take into consideration the SESR criteria. | allowable ESR. Hence the maximum permissible packet loss probability | |||
is 4%. Of course, this estimate is extremely optimistic, and | ||||
furthermore does not take into consideration the SESR criteria. | ||||
As previously explained, a SES is declared whenever AIS is detected. | As previously explained, a SES is declared whenever AIS is detected. | |||
There is a major difference between structure-aware and structure- | There is a major difference between structure-aware and structure- | |||
agnostic transport in this regards. When a packet is lost SAToP | agnostic transport in this regards. When a packet is lost SAToP | |||
outputs an "all-ones" pattern to the TDM circuit, which is | outputs an "all-ones" pattern to the TDM circuit, which is | |||
interpreted as AIS according to G.775 [G775]. For E1 circuits, G.775 | interpreted as AIS according to G.775 [G775]. For E1 circuits, G.775 | |||
specifies for AIS to be detected when four consecutive TDM frames | specifies that AIS is detected when four consecutive TDM frames have | |||
have no more than 2 alternations. This means that if a PW packet or | no more than 2 alternations. This means that if a PW packet or | |||
consecutive packets containing at least four frames are lost, and | consecutive packets containing at least four frames are lost, and | |||
four or more frames of "all-ones" output to the TDM circuit, a SES | four or more frames of "all-ones" output to the TDM circuit, a SES | |||
will be declared. Thus burst packet loss, or packets containing a | will be declared. Thus burst packet loss, or packets containing a | |||
large number of TDM frames, lead SAToP to cause high SESR, which is | large number of TDM frames, lead SAToP to cause high SESR, which is | |||
20 times more restricted than ESR. On the other hand, since | 20 times more restricted than ESR. On the other hand, since | |||
structure-aware transport regenerates the correct frame alignment | structure-aware transport regenerates the correct frame alignment | |||
pattern, even when the corresponding packet has been lost, packet | pattern, even when the corresponding packet has been lost, packet | |||
loss will not cause declaration of SES. This is the main reason that | loss will not cause declaration of SES. This is the main reason that | |||
SAToP is much more vulnerable to packet loss than the structure-aware | SAToP is much more vulnerable to packet loss than the structure-aware | |||
methods. | methods. | |||
skipping to change at page 19, line 41 | skipping to change at page 19, line 44 | |||
will be intermediate between the extremely pessimistic estimates and | will be intermediate between the extremely pessimistic estimates and | |||
the extremely optimistic ones. In order to numerically gauge the | the extremely optimistic ones. In order to numerically gauge the | |||
situation, we have modeled the network as a four-state Markov model, | situation, we have modeled the network as a four-state Markov model, | |||
(corresponding to a successfully received packet, a packet received | (corresponding to a successfully received packet, a packet received | |||
within a loss burst, a packet lost within a burst, and a packet lost | within a loss burst, a packet lost within a burst, and a packet lost | |||
when not within a burst). This model is an extension of the widely | when not within a burst). This model is an extension of the widely | |||
used Gilbert model. We set the transition probabilities in order to | used Gilbert model. We set the transition probabilities in order to | |||
roughly correspond to anecdotal evidence, namely low background | roughly correspond to anecdotal evidence, namely low background | |||
isolated packet loss, and infrequent bursts wherein most packets are | isolated packet loss, and infrequent bursts wherein most packets are | |||
lost. Such simulation shows that up to 0.5% average packet loss may | lost. Such simulation shows that up to 0.5% average packet loss may | |||
occur and the recovered TDM still conform to the G.826 ESR and SESR | occur and the recovered TDM still conforms to the G.826 ESR and SESR | |||
criteria. | criteria. | |||
Appendix B. Effect of Packet Loss on Voice Quality for TDM PWs | Appendix B. Effect of Packet Loss on Voice Quality for TDM PWs | |||
Packet loss in voice traffic can cause in gaps or artifacts that | Packet loss in voice traffic can cause gaps or artifacts that result | |||
result in choppy, annoying or even unintelligible speech. The | in choppy, annoying or even unintelligible speech. The precise | |||
precise effect of packet loss on voice quality has been the subject | effect of packet loss on voice quality has been the subject of | |||
of detailed study in the VoIP community, but VoIP results are not | detailed study in the VoIP community, but VoIP results are not | |||
directly applicable to TDM PWs. This is because VoIP packets | directly applicable to TDM PWs. This is because VoIP packets | |||
typically contain over 10 milliseconds of the speech signal, while | typically contain over 10 milliseconds of the speech signal, while | |||
multichannel TDM packets may contain only a single sample, or perhaps | multichannel TDM packets may contain only a single sample, or perhaps | |||
a very small number of samples. | a very small number of samples. | |||
The effect of packet loss on TDM PWs has been previously reported | The effect of packet loss on TDM PWs has been previously reported | |||
[I-D.stein-pwe3-tdm-packetloss]. In that study it was assumed that | [I-D.stein-pwe3-tdm-packetloss]. In that study it was assumed that | |||
each packet carried a single sample of each TDM timeslot (although | each packet carried a single sample of each TDM timeslot (although | |||
the extension to multiple samples is relatively straightforward and | the extension to multiple samples is relatively straightforward and | |||
does not drastically change the results). Four sample replacement | does not drastically change the results). Four sample replacement | |||
skipping to change at page 20, line 40 | skipping to change at page 20, line 43 | |||
The four algorithms were compared in a controlled experiment in which | The four algorithms were compared in a controlled experiment in which | |||
speech data was selected from English and American English subsets of | speech data was selected from English and American English subsets of | |||
the ITU-T P.50 Appendix 1 corpus [P.50App1] and consisted of 16 | the ITU-T P.50 Appendix 1 corpus [P.50App1] and consisted of 16 | |||
speakers, eight male and eight female. Each speaker spoke either | speakers, eight male and eight female. Each speaker spoke either | |||
three or four sentences, for a total of between seven and 15 seconds. | three or four sentences, for a total of between seven and 15 seconds. | |||
The selected files were filtered to telephony quality using modified | The selected files were filtered to telephony quality using modified | |||
IRS filtering and downsampled to 8 KHz. Packet loss of 0, 0.25, 0.5, | IRS filtering and downsampled to 8 KHz. Packet loss of 0, 0.25, 0.5, | |||
0.75, 1, 2, 3, 4 and 5 percent were simulated using a uniform random | 0.75, 1, 2, 3, 4 and 5 percent were simulated using a uniform random | |||
number generator (bursty packet loss was also simulated but is not | number generator (bursty packet loss was also simulated but is not | |||
reported here). For each file the four methods of lost sample | reported here, given it is not the worst-case). For each file the | |||
replacement were applied and the Mean Opinion Score (MOS) was | four methods of lost sample replacement were applied and the Mean | |||
estimated using PESQ [P862]. Figure 5 depicts the PESQ-derived MOS | Opinion Score (MOS) was estimated using PESQ [P862]. Figure 9 | |||
for each of the four replacement methods for packet drop | depicts the PESQ-derived MOS for each of the four replacement methods | |||
probabilities up to 5%. | for packet drop probabilities up to 5%. | |||
-------------------------------------------------------------------- | -------------------------------------------------------------------- | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I PESQ-MOS as a function of packet drop probability I | I PESQ-MOS as a function of packet drop probability I | |||
I I | I I | |||
skipping to change at page 21, line 27 | skipping to change at page 21, line 27 | |||
I (only in PDF version) I | I (only in PDF version) I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
I I | I I | |||
-------------------------------------------------------------------- | -------------------------------------------------------------------- | |||
Figure 5 PESQ derived MOS as a function of packet drop probability | Figure 9 PESQ derived MOS as a function of packet drop probability | |||
For all cases the MOS resulting from the use of zero insertion is | For all cases the MOS resulting from the use of zero insertion is | |||
less than that obtained by replacing with the previous sample, which | less than that obtained by replacing with the previous sample, which | |||
in turn is less than that of linear interpolation, which is slightly | in turn is less than that of linear interpolation, which is slightly | |||
less than that obtained by statistical interpolation. | less than that obtained by statistical interpolation. | |||
Unlike the artifacts speech compression methods may produce when | Unlike the artifacts speech compression methods may produce when | |||
subject to buffer loss, packet loss here effectively produces | subject to buffer loss, packet loss here effectively produces | |||
additive white impulse noise. The subjective impression is that of | additive white impulse noise. The subjective impression is that of | |||
static noise on AM radio stations or crackling on old phonograph | static noise on AM radio stations or crackling on old phonograph | |||
End of changes. 39 change blocks. | ||||
137 lines changed or deleted | 151 lines changed or added | |||
This html diff was produced by rfcdiff 1.41. The latest version is available from http://tools.ietf.org/tools/rfcdiff/ |