Aggregate traffic performance with active queue management and drop from tail
Download
1 / 23

Aggregate Traffic Performance with Active Queue Management and Drop from Tail - PowerPoint PPT Presentation


  • 126 Views
  • Uploaded on

Aggregate Traffic Performance with Active Queue Management and Drop from Tail. Christophe Diot, Gianluca Iannaccone, Martin May Sprint ATL, Universit à di Pisa, Activia www.sprintlabs.com. Active Queue Management. queue. average. instantaneous. drop size. function. sharp. RED.

loader
I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
capcha
Download Presentation

PowerPoint Slideshow about ' Aggregate Traffic Performance with Active Queue Management and Drop from Tail' - jenny


An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
Aggregate traffic performance with active queue management and drop from tail

Aggregate Traffic Performance with Active Queue Management and Drop from Tail

Christophe Diot, Gianluca Iannaccone, Martin May

Sprint ATL, Universitàdi Pisa, Activia

www.sprintlabs.com


Active queue management
Active Queue Management and Drop from Tail

queue

average

instantaneous

drop size

function

sharp

RED

Drop from Tail

smooth

Gentle

Gentle RED

RED

Instantaneous


Original red
Original RED and Drop from Tail

1

max-p

min-tresh

max-tresh

buffer size


Revised red
Revised RED and Drop from Tail

1

max-p

min-tresh

max-tresh

buffer size


Gentle red
Gentle RED and Drop from Tail

1

max-p

min-tresh

max-tresh

buffer size

2 max-tresh


Experiments

Testbed and Drop from Tail

with CISCO routers (7500)

with Dummynet

We use “recommended” RED and GRED parameters

Heterogeneous delays (120 to 180 ms)

Experiments


Traffic characteristics

16 to and Drop from Tail256 TCP connections sharing the bottleneck.

Experimental traffic generated by Chariot

long-lived TCP connections.

more “realistic” traffic mix:

90% short lived TCP connections (up to 20 packets)

10 % long lived TCP connections

1Mbps UDP in both cases

Traffic characteristics


Testbed cisco routers
Testbed (CISCO routers) and Drop from Tail

7500

7500

10 Megs


Testbed dummynet
Testbed (Dummynet) and Drop from Tail

7500

7500

10 Megs

Dummy

net

100 Megs


What is dummynet
What is Dummynet? and Drop from Tail

application

dummynet

network


Metrics observed

Aggregate goodput through a router and Drop from Tail

TCP and UDP loss rate

Consecutive losses

Queuing behavior

Metrics observed









In summary

No significant difference on goodput, TCP losses and UDP losses.

On consecutive losses, clear advantage to GRED and GRED-I.

“gentle” modification solves many RED problems.

Oscillations: no clear winner. Traffic seems to be the determining factor.

In summary ...


From the isp standpoint

Not clear there is an advantage in deploying RED, GRED, or GRED-I.

Maybe GRED-I is an option if one can find a “universal” exponential dropping function.

ECN will work with any scheme.

Not clear the solution is in the AQM space.

From the ISP standpoint ...



About fair queuing

Not only feasible … easy at the edges! GRED-I.

www.agere.com (an example)

vendors support from 64k to 200k flows

Really fair

everybody gets what he/she paid for

local signaling (end host to CPE)

About Fair Queuing ...



ad