Measurements of Internet performance for NIIT, Pakistan Jan – Feb 2004 - PowerPoint PPT Presentation

Measurements of internet performance for niit pakistan jan feb 2004 l.jpg
Download
1 / 12

PingER. Measurements of Internet performance for NIIT, Pakistan Jan – Feb 2004. From Les Cottrell, SLAC For presentation by Prof. Arshad Ali, NIIT. PingER. Results: Worldwide performance. Performance is improving Developed world improving factor of 10 in 4-5 years

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

Download Presentation

Measurements of Internet performance for NIIT, Pakistan Jan – Feb 2004

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


Measurements of internet performance for niit pakistan jan feb 2004 l.jpg

PingER

Measurements of Internet performance for NIIT, Pakistan Jan – Feb 2004

From Les Cottrell, SLAC

For presentation by Prof. Arshad Ali, NIIT

PingER


Results worldwide performance l.jpg

Results: Worldwide performance

  • Performance is improving

  • Developed world improving factor of 10 in 4-5 years

  • S.E. Europe, C.AsiaRussia, catching up

  • India & Africa worse off & falling behind

  • Developing world 3-10 years behind

  • Many institutes in developing world have less performance than a household in N. America or Europe!!


To pakistan performance l.jpg

To Pakistan performance

Routes: ESnet (hops 3-8) - DC

ATT (9-21) - Karachi

Karachi

NIIT/Rawalpindi

Routes: ESnet (hops 3-6) - SNV

SINGTEL (7-12) - Karachi

Pakistan Telecom

Karachi

Rawalpindi

Loss %

Islamabad

RTT ms

Routes: ESnet (hops 3-6) - SNV

SINGTEL (7-12) - Karachi

Pakistan Telecom

Karachi

Lahore

Lahore


From pakistan performance l.jpg

From Pakistan Performance

Route:Pakistan Telecom (2-5)

Rawalpindi, Karachi

SingTel (6-10)

ESnet (11-14) - PAIX

NIIT to SLAC

Route: Pakistan Telecom (2-5) Rawalpindi, KarachiConcert (6-9) LondonDataTAG (11-12) .de

NIIT to CERN

Note similarities, probably due to common bottleneck, probably in Pakistan


Niit performance from u s slac l.jpg

NIIT performance from U.S. (SLAC)

Preliminary results, started measurements end Dec 2003.

Nb. Heavy losses during congested day-times

Avg daily:

loss~1-2%,

RTT~320ms

Ping RTT & Loss

Bandwidth measurements using packet pair dispersion & TCP (Jan 2004)

abing (pkt-pair dispersion):Average To NIIT: ~350Kbits/s From NIIT: ~365 Kbits/s

Iperf/TCP (with SLAC): Average: To NIIT: ~320Kbits/s From NIIT: ~330Kbits/s

Iperf/TCP (with CERN): Average: To NIIT: ~270Kbits/s From NIIT: ~300Kbits/s

Can also derive throughput (assuming standard TCP) from RTT & loss (monthly) using: BW~1.2*S(1460B)/(RTT*sqrt(loss)  ~ 260Kbits/s (SLAC to NIIT)

~ 630Kbits/s (NIIT to SLAC | CERN)

Nominal path bottleneck capacity 364 Kbits/s


Available bandwidth feb 04 after upgrade l.jpg

Available Bandwidth (Feb ‘04 after upgrade)

  • green line is the bandwidth capacity of current bottleneck

    • deduced from the minimum packet separation

  • blue line is available bandwidth = capacity-cross-traffic.

  • Use available bandwidth estimator (abing)

    • Uses packet pair dispersion

    • Low impact, 40*1450Byte packets

    • Repeat once/minute

    • Client at SLAC, mirror/server at NIIT

  • Iperf confirms with:

    • 948Kbps (2streams),

    • 952Kbps (4streams),

    • 1042Kbps (10streams)


To ncp pakistan l.jpg

To NCP Pakistan

  • Cannot use PingER to measure to ncp.edu.pk

    • Pings blocked at FLAG router (62.216.145.154, AS15412) on way to Comsats (Pakistani ISP)

  • Working with NCP to try and resolve

    • Trying to contact FLAG

  • Using abing instead

    • Indicates 2Mbps

    • But link is 384Kbps

  • Iperf shows 235 - 245 Kbps

  • Rate limiting or shaping?

~ 2MBits/s, but link is 384Kbps

Looking for discrepancy


Within pakistan l.jpg

Within Pakistan

  • SLAC – Karachi U:

    • ESnet (hops 3-8) – DC ATT (9-21) – Karachi

  • SLAC – NIIT RawalpindiI:

    • ESnet (hops 3-6) – SNV, SINGTEL (7-12) – Karachi, Pakistan Telecom Karachi-Rawalpindi

  • SLAC - U Lahore, similar to NIIT

  • SLAC – NSC:

    • ESnet (hops 1-6), C&W (7-11) Santa Clara – NY, FLAG (12-16) NY – London – Karlsruhe, Comsats

  • NIIT – NSC (Rawalpindi – Islamabad) few miles apart,

    • No peering in Pakistan, can this be changed?

    • Route goes via England:

      • PIE (hops 1-5), Concert (6-9)- London, FLAG (10-14) London – Karachi, Comsats (15)

    • Takes longer than to SLAC


Conclusions l.jpg

Conclusions

  • Big performance differences to sites, depend on ISP (at least 3 ISPs seen for Pakistan A&R sites)

  • To NIIT:

    • Before upgrade got about 300Kbps - 380Kbps at best

    • After upgrade get 1 Mbps, as expected

    • The bottleneck appears to be in Pakistan

    • There is often congestion (packet loss & extended RTTs) during busy periods each weekday

    • Video will probably be sensitive to packet loss, so it may depend on the time of day

    • H.323 (typically needs 384Kbps + 64Kbps), would appear to have been be marginal at best before upgrade, since upgrade has been very successful.

  • No peering Pakistan between NIIT and NSC


Bulk data transfer l.jpg

Bulk Data Transfer

  • Transfer time to send a file of various sizes between 2 sites with given capacity

    • assume can utilize 50% of capacity

    • format hours:mins:seconds

File size

Typical BaBar file sizes 500MB-1GB

PingER


Interactive use l.jpg

Interactive Use

  • Voice needs RTT < 250ms or else listener does not know when to speak

  • RTT > 400ms makes productive interactive work such as interactive telnet/X-windows style typing difficult

    • Screen does not match the keyboard, especially when correcting text

  • Losses:

    • Losses > 10% TCP connections fail

    • Losses >4-6% make video conferencing unintelligible for non-native language speakers

    • Losses of > 3-5% make TCP perform badly

    • Random loss of 2.5% will make Voice over IP annoying every 30 seconds or so

    • More realistic burst losses will cause VoIP to be annoying at >1% losses

PingER


More information l.jpg

More information

  • NUST Institute of Information Technology (NIIT)

    • http://www.niit.edu.pk/

  • PingER project

    • http://www-iepm.slac.stanford.edu/pinger/

  • ABwE available bandwidth estimator

    • www-iepm.slac.stanford.edu/bw/abwe/abwe-cf-iperf.html


  • Login