Blue (queue management algorithm): Difference between revisions

Revision as of 13:49, 3 April 2011

Latest revision as of 02:15, 9 March 2025

(42 intermediate revisions by 25 users not shown)

Line 1:

'''<span style="font-variant: small-caps;">Blue</span>''' is a [[scheduling discipline]] for the [[network scheduler]] developed by graduate student Wu-chang Feng for Professor [[Kang G. Shin]] at the [[University of Michigan]] and others at the [[Thomas J. Watson Research Center]] of [[IBM]] in 1999.<ref name="mich">{{CitationCite journal |title=BLUE: A New Class of Active Queue Management Algorithms |author1=Wu-chang Feng |author2=Dilip D. Kandlur |author3=Debanjan Saha |author4=Kang G. Shin |yeardate=April 1999 |month=April |url=http://www.eecs.umich.edu/techreports/cse/99/CSE-TR-387-99.pdf |journalpublisher=U. University of Michigan |journal= Computer Science Technical Report |issue=CSE-TR-387-99CSE–TR–387–99 |accessdate=2010-12-22 June 8, 2013 }}</ref> is an [[Active Queue Management]] algorithm. Like [[Random early detection|RED]], it operates by randomly dropping or [[Explicit Congestion Notification|ECN]]-marking packets in a router's queue before it overflows. Unlike RED, however, it requires little or no tuning on the part of the network administrator.

==Functioning==

==Operation of <span style="font-variant: small-caps;">B'''lue'''</span>== <!--Bolding small cap letters looks MUCH better on Firefox; don't know about other browsers.-->

ALike <span[[random style="font-variant:early detection]] (RED), Blue operates by randomly dropping or marking packet with [[explicit congestion notification]] mark before the transmit buffer of the [[network interface controller]] overflows. Unlike RED, however, it requires little or no tuning to be performed by the network administrator. A small-caps;">Blue</span> queue maintains a drop/mark probability ''p'', and drops/marks packets with probability ''p'' as they enter the queue. Whenever the queue overflows, ''p'' is increased by a small constant ''p_di'', and whenever the queue is empty, ''p'' is decreased by a constant ''p_id&nbsp;< p_di''.

AssumingIf the mix of traffic on the interface doesn'tdoes not change, ''p'' will slowly converge to a value that keeps the queue within its bounds with full link utilisationutilization.

A <span style="font-variant: small-caps;">Blue</span> queue maintains a drop/mark probability ''p'', and drops/marks packets with probability ''p'' as they enter the queue. Whenever the queue overflows, ''p'' is increased by a small constant ''p_d'', and whenever the queue is empty, ''p'' is decreased by a constant ''p_i<p_d''.

===Stochastic fair Blue===

Assuming the mix of traffic on the interface doesn't change, ''p'' will slowly converge to a value that keeps the queue within its bounds with full link utilisation.

The main flaw of <span style="font-variant: small-caps;">Blue</span>, which it shares with most single-queue queueing[[queuing disciplinesdiscipline]]s, is that it doesn'tdoes not distinguish between [[Traffic flow (computer networking)|traffic flows]], andbut treats all flows as a single aggregate. Therefore, a single aggressive flow can push packets out of the queue packets belonging to other, better behaved, flows.

Stochastic fair Blue (SFB) is a stochastically fair variant of Blue which hashes flows and maintains a different mark/drop probability for each hash value. Assuming no hash collisions, SFB is able to provide a fair share of buffer space for every flow. In the presence of hash collisions, SFB is only stochastically fair.<ref>{{Cite book|author1=Wu-Chang Feng |author2=Dilip D. Kandlur |author3=Debanjan Saha |author4=Kang G. Shin |title=Proceedings IEEE INFOCOM 2001. Conference on Computer Communications. Twentieth Annual Joint Conference of the IEEE Computer and Communications Society (Cat. No.01CH37213) |chapter=Stochastic fair blue: A queue management algorithm for enforcing fairness |url=http://www.thefengs.com/wuchang/blue/41_2.PDF |date=April 2001 |pages=1520–1529 |doi=10.1109/INFCOM.2001.916648 |accessdate= June 8, 2013 |volume=3|isbn=978-0-7803-7016-6 |citeseerx=10.1.1.11.4235 |s2cid=5902623 }}</ref>

==Stochastic Fair <span style="font-variant: small-caps;">B'''lue'''</span>==

Unlike other stochastically fair queuing disciplines, such as SFQ ([[Stochastic Fairness Queuing]]), SFB can be implemented using a [[Bloombloom filter]] rather than a [[hash table]], which dramatically reduces its storage requirements when the number of flows is large.

The main flaw of <span style="font-variant: small-caps;">Blue</span>, which it shares with most single-queue queueing disciplines, is that it doesn't distinguish between flows, and treats all flows as a single aggregate. Therefore, a single aggressive flow can push out of the queue packets belonging to other, better behaved, flows.

When a flow's drop/mark probability reaches 1, the flow has been shown to not react to congestion indications from the network. Such an inelastic flow is put in a "[[penalty box]]", and rate-limited.

===Resilient stochastic fair Blue===

Stochastic Fair <span style="font-variant: small-caps;">Blue</span> (SFB)<ref>{{Citation |author1=Wu-Chang Feng |author2=Dilip D. Kandlur |author3=Debanjan Saha |author4=Kang G. Shin |title=Stochastic Fair Blue: an algorithm for enforcing fairness |url=http://www.thefengs.com/wuchang/blue/41_2.PDF |journal=Proc. INFOCOM 2001 |year=2001 |month=April |pages=1520–1529 |doi=10.1109/INFCOM.2001.916648 |accessdate=2010-01-02}}</ref> is a stochastically fair variant of <span style="font-variant: small-caps;">Blue</span> which hashes flows and maintains a different mark/drop probability for each hash value. Assuming no hash collisions, SFB is able to provide a fair share of buffer space for every flow. In the presence of hash collisions, SFB is only stochastically fair.

Many scheduling algorithms, including the fairness-aimed ones, are notably vulnerable to spoofing [[distributed denial-of-service]] (DDoS) attacks. A resilient stochastic fair Blue (RSFB) algorithm was proposed in 2009 against spoofing DDoS attacks. The basic idea behind RSFB is to record the responsive normal TCP flows and rescue their dropped packets. RSFB algorithm is effective in preserving the TCP throughput in the presence of spoofing DDoS attacks.<ref name=RSFB>{{Cite book |author1=Changwang Zhang |author2=Jianping Yin |author3=Zhiping Cai |name-list-style=amp |url= http://sites.google.com/site/cwzhangres/home/files/RSFBaResilientStochasticFairBluealgorithmagainstspoofingDDoSattacks.pdf |title= RSFB: a Resilient Stochastic Fair Blue algorithm against spoofing DDoS attacks |journal= International Symposium on Communication and Information Technology |year= 2009 |pages= 1566–1567 |isbn= 978-1-4244-4521-9 |accessdate= June 8, 2013 }} [http://portal.acm.org/citation.cfm?id=1789954.1790341 Abstract]</ref>

==Implementations==

Unlike other stochastically fair queuing disciplines, such as SFQ, SFB can be implemented using a [[Bloom filter]] rather than a hash table, which dramatically reduces its storage requirements when the number of flows is large.

An implementation of Blue is part of [[ALTQ]], the [[network scheduler]] for BSD Unix.<ref>{{Cite web |title= Blue |work= Web page |author= Wu-chang Feng |url= http://www.thefengs.com/wuchang/blue/ |accessdate= June 8, 2013 }}</ref>

An implementation of SFB for [[Linux]] was included in the [[Linux kernel]] in version 2.6.39.<ref>[http://kernelnewbies.org/Linux_2_6_39#head-87ffd4407af29460251c521e0228fe0ac9219d4b Kernel Newbies - Linux 2.6.39 - Networking]</ref><ref>{{cite web |url=https://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/tree/net/sched/sch_sfb.c |title=SFB Linux kernel network scheduler module |publisher=[[kernel.org]] |accessdate=2013-09-07}}</ref><ref>{{Cite web |title= Stochastic Fair Blue for the Linux kernel |author= Juliusz Chroboczek |url= http://www.pps.univ-paris-diderot.fr/~jch/software/sfb/ |accessdate= June 8, 2013 }}</ref>

When a flow's drop/mark probability reaches 1, the flow has been shown to not react to congestion indications from the network. Such an inelastic flow is put in a "[[penalty box]]", and rate-limited.

==Implementations==

==References==

{{Reflist}}

<references/>

==External links==

* [http://www.thefengs.com/wuchang/blue/ Wu-chang Feng's page about Blue and SFB].

* [http://www.pps.jussieu.fr/~jch/software/sfb/ An implementation of SFB for the Linux kernel].

[[Category:Packets (information technology)]]