23
JIVE VLBI Network Meeting 28 Jan 2004 R. Hughes-Jones Manchester Brief Report on Tests Related to the e-VLBI Project Richard Hughes-Jones The University of Manchester DataGrid WP7 – Dante Tests on the GÉANT Core End-2-End Measurements from the 4 th Year VLBI Project at Manchester DiskPack-2-Memory Throughput and PCI Activity in a Mark5 PC Update on 1 and 10 Gigabit Ethernet NICs in the PC

JIVE VLBI Network Meeting 28 Jan 2004 R. Hughes-Jones Manchester Brief Report on Tests Related to the e-VLBI Project Richard Hughes-Jones The University

  • View
    215

  • Download
    0

Embed Size (px)

Citation preview

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Brief Report on Tests Related to the e-VLBI Project

Richard Hughes-JonesThe University of Manchester

DataGrid WP7 – Dante Tests on the GÉANT CoreEnd-2-End Measurements from the 4th Year VLBI Project at ManchesterDiskPack-2-Memory Throughput and PCI Activity in a Mark5 PCUpdate on 1 and 10 Gigabit Ethernet NICs in the PC

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

DataGrid WP7 – Dante Tests on the GÉANT Core Set-up

Supermicro PC in: London GEANT PoP Amsterdam GEANT PoP

Smartbits in: London GEANT PoP Frankfurt GEANT PoP

Long link UK-SE-DE2-IT-CH-FR-BE-NL

Short Link UK-FR-BE-NL

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Tests GÉANT Core: UDP throughput UDP Throughput London-Amsterdam Available BW to packet on wire Then 1/t Wire rate 998 Mbit/s

for packets > 1400 bytes

Packet Loss None for large packets

Dips in BW lined to packet loss SysKonnect NIC int. per packet CPU load important

uk-nl_20tg4-hs-w100_01Oct03

0

100

200

300

400

500

600

700

800

900

1000

0 5 10 15 20 25 30 35 40Spacing between frames us

Rec

v W

ire r

ate

Mbi

ts/s

50 bytes 100 bytes 200 bytes 400 bytes 600 bytes 800 bytes 1000 bytes 1200 bytes 1400 bytes 1472 bytes

0102030405060708090

100

0 5 10 15 20 25 30 35 40Spacing between frames us

% P

acke

t los

s

50 bytes 100 bytes 200 bytes 400 bytes 600 bytes 800 bytes 1000 bytes 1200 bytes 1400 bytes 1472 bytes

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Tests GÉANT Core: Packet re-ordering Effect of Packet size London-Amsterdam Packets at 10 µs – line speed 10,000 sent Packet Loss ~ 0.1%

Re-order Distribution

Packet re-order uk-nl 10,000 BE sent wait 10 us 01 Oct 03

0

5

10

15

20

25

30

0 500 1000 1500Packet size bytes

Out

of o

rder

% 0

10

20

30

1400 1401 1402 1403 1404Packet size bytes

Ou

t o

f o

rde

r %

Packet re-order uk-nl 10,000 sent wait 10 us

0

100

200

300

400

500

0 1 2 3 4 5 6 7 8 9Length out-of-order

No.

Pac

kets 1400 bytes

1401 bytes

1402 bytes

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Tests GÉANT Core: Packet re-ordering Effect of LBE background Amsterdam-London BE Test flow Packets at 10 µs – line speed 10,000 sent Packet Loss ~ 0.1%

Re-order Distributions:

UDP 1472 bytes NL-UK-lbexxx_7nov03

02468

101214161820

2 2.2 2.4 2.6 2.8 3 3.2Total Offered Rate Gbit/s

% O

ut o

f ord

er

hstcpStandard TCP line speed90% line speed

Packet re-order 1472 bytes uk-nl 21 Oct 03 10,000 sent wait 10 us

020000400006000080000

100000120000140000160000180000200000

1 2 3 4 5 6 7 8 9Length out-of-order

No.

Pac

kets

0 % lbe

10 % lbe

20 % lbe

30 % lbe

40 % lbe

50 % lbe

60 % lbe

70 % lbe

80 % lbe

Packet re-order 1400 bytes uk-nl 21 Oct 03 10,000 sent wait 10 us

0500

100015002000250030003500400045005000

1 2 3 4 5 6 7 8 9Length out-of-order

No.

Pac

kets

0 % lbe

10 % lbe

20 % lbe

30 % lbe

40 % lbe

50 % lbe

60 % lbe

70 % lbe

80 % lbe

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

VLBI Project: Test Topology

SuperJANET4

Jodrell

Manchester

SURFnet

JIVEDwingaloo

Adam MathewsSteve O’TooleUniv of Manchester

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Gnt5-DwMk5 11Nov03-1472 bytes

0

2

4

6

8

10

12

0 5 10 15 20 25 30 35 40Spacing between frames us

% P

acket

loss

Gnt5-DwMk5

DwMk5-Gnt5

Manchester to Dwingeloo 2.0G Hz Xeon 1.2 GHz PIII

Re-ordering vs Offered Load

VLBI Project: Throughput

Gnt5-DwMk5 11Nov03/DwMk5-Gnt5 13Nov03-1472bytes

0

200

400

600

800

1000

1200

0 5 10 15 20 25 30 35 40Spacing between frames us

Recv W

ire r

ate

Mbits/s

Gnt5-DwMk5

DwMk5-Gnt5

Gnt5-DwMk5 11Nov03 1472 bytes

020406080

100

0 5 10 15 20 25 30 35 40Spacing between frames us

% K

erne

l S

ende

r

Gnt5-DwMk5 11Nov03 1472 bytes

020406080

100

0 5 10 15 20 25 30 35 40Spacing between frames us

% K

erne

l R

ecei

ver

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

1472 byte Packets man -> JIVE FWHM 22 µs (B2B 3 µs )

VLBI Project: Jitter & 1-way Delay

1472 bytes w=50 jitter Gnt5-DwMk5 28Oct03

0

2000

4000

6000

8000

10000

0 20 40 60 80 100 120 140

Jitter us

N(t

)

1472 bytes w=50 jitter Gnt5-DwMk5 28Oct03

1

10

100

1000

10000

0 20 40 60 80 100 120 140

Jitter us

N(t

)

1472 bytes w12 Gnt5-DwMk5 21Oct03

0

2000

4000

6000

8000

10000

12000

2000 2100 2200 2300 2400 2500 2600 2700 2800 2900 3000Packet No.

1-w

ay

de

lay

us

1472 bytes w12 Gnt5-DwMk5 21Oct03

0

2000

4000

6000

8000

10000

12000

0 1000 2000 3000 4000 5000Packet No.

1-w

ay d

elay

us

1-way Delay – note the packet loss (points with 0 –way delay)

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Measure the time between lost packets in the time series of packets sent.

Lost 1410 in 0.6s Is it a Poisson process? Assume Poisson is

stationary λ(t) = λ

Use Prob. Density Function:

P(t) = λ e-λt

Mean λ = 2360 / s[426 µs]

Plot log: slope -0.0028expect -0.0024

Could be additional process involved

VLBI Project: Packet Loss Distributionpacket loss distribution 12b bin=12us

0

10

20

30

40

50

60

70

80

12 72 132

192

252

312

372

432

492

552

612

672

732

792

852

912

972

Time between lost frames (us)

Num

ber

in B

in

Measured

Poisson

packet loss distribution 12b

y = 41.832e-0.0028x

y = 39.762e-0.0024x

1

10

100

0 500 1000 1500 2000

Time between frames (us)

Num

ber

in B

in

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Divide time series of packets into 1000 slices of 50 packets

Total lost packets 1410 Average number / slice = 1.4

Calc Poisson Probability P(n, µ) = µ n e -µ

n!

Curves close but not exact Could be more than 1 process

VLBI Project: Packet Loss – Is it Poisson?

0

50

100

150

200

250

300

350

400

0 5 10 15n num lost in sub-sample

N(n

)

run12b

1

1.3

1.4

1.8

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Aggregated Variance Method Divide time series length N into

blocks of size m Calc mean of each section Xm(k)

k= 1 … N/m Calc variance VXm of these Xm(k) Vary m size of the blocks

Plot on log-log & fit slope β Hurst parameter H

β = 2H -2 Measure:

β = -0.355 which gives H 0.822 H =1 no long range dependence

VLBI Project: Packet Loss – Long Range Effects?

y = -0.355x + 2.8826

0

0.5

1

1.5

2

2.5

3

3.5

4

0 0.5 1 1.5 2 2.5 3sub-sample size Log10( m )

Ag

gri

ga

te-v

ari

an

ce L

og

10

( X

(m)

)

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Traffic Flows Manchester – NetNorthWest - SuperJANET Access links

Two 1 Gbit/s

Access links:SJ4 to GÉANT GÉANT to SurfNet

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Throughput & PCI transactions on the Mark5 PC:

Read /Write

n bytes

Wait timetime

Mark5 uses Supermicro P3TDLE 1.2 GHz PIII Mem bus 133/100 MHz 2 *64bit 66 MHz PCI 4 32bit 33 MHz PCI

SuperStor

NIC

Input Card

IDEDiscPack

Ethernet

Logic AnalyserDisplay

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

PCI Activity: Read 1 data block 0 wait time

Data block 131,072 bytes Read consists of:

Setup CSRs Data movement Update CSRs

Data block contains PCI bursts 4096 bytes long

For 0 wait between reads: Data block 608µs long Then 655µs gap

PCI transfer rate 830 Mbit/s Read_sstor rate 778 Mbit/s

(97 Mbyte/s)

CSR Access

PCI Burst

Data Block

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

PCI Activity: Read Multiple data blocks 0 wait Read 999424 bytes Each Data block:

Setup CSRs Data movement Update CSRs

For 0 wait between reads: Data blocks ~600µs long

take ~6 ms Then 744µs gap

PCI transfer rate 1188Mbit/s(148.5 Mbytes/s)

Read_sstor rate 778 Mbit/s (97 Mbyte/s)

PCI bus occupancy: 68.44% Concern about Ethernet Traffic

64 bit 33 MHz PCI needs ~ 82% for 930 Mbit/s Expect ~360 Mbit/s

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

PCI Activity: Read Throughput

Flat then 1/t dependance ~ 860 Mbit/s for Read blocks

>= 262144 bytes

CPU load ~20% Concern about CPU load

needed to drive Gigabit link

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

10 GigEthernet: Throughput 1500 byte MTU gives ~ 2 Gbit/s Used 16144 byte MTU max user length 16080 DataTAG Supermicro PCs Dual 2.2 GHz Xenon CPU FSB 400 MHz PCI-X mmrbc 512 bytes wire rate throughput of 2.9 Gbit/s

CERN OpenLab HP Itanium PCs Dual 1.0 GHz 64 bit Itanium CPU FSB 400 MHz PCI-X mmrbc 512 bytes wire rate of 5.7 Gbit/s

SLAC Dell PCs giving a Dual 3.0 GHz Xenon CPU FSB 533 MHz PCI-X mmrbc 4096 bytes wire rate of 5.4 Gbit/s

an-al 10GE Xsum 512kbuf MTU16114 27Oct03

0

1000

2000

3000

4000

5000

6000

0 5 10 15 20 25 30 35 40Spacing between frames us

Rec

v W

ire

rate

Mb

its/

s

16080 bytes 14000 bytes 12000 bytes 10000 bytes 9000 bytes 8000 bytes 7000 bytes 6000 bytes 5000 bytes 4000 bytes 3000 bytes 2000 bytes 1472 bytes

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

10 GigEthernet: PCI-X Activity

Supermicro P4DP8-2G motherboard

Packet transmissionMemory to NIC

Packet receptionNIC to Memory

CSR Access

Interrupt

PCI-X segment 512 bytes

Transfer of 16114 bytes

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

10 GigEthernet: Tuning PCI-X

mmrbc 1024 bytes

mmrbc 2048 bytes

mmrbc 4096 bytes

Transfer of 16114 bytes

mmrbc 512 bytes

PCI-X segment 2048 bytes

CSR Access

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

10 GigEthernet at SC2003 BW Challenge (Phoenix) Three Server systems with 10 GigEthernet NICs Used the DataTAG altAIMD stack 9000 byte MTU Streams From SLAC/FNAL booth in Phoenix to:

Pal Alto PAIX

Chicago Starlight Amsterdam SARA

10 Gbits/s throughput from SC2003 to PAIX

0

1

2

3

4

5

6

7

8

9

10

11/19/0315:59

11/19/0316:13

11/19/0316:27

11/19/0316:42

11/19/0316:56

11/19/0317:11

11/19/0317:25 Date & Time

Throughput

Gbits/s

Router to LA/PAIXPhoenix-PAIX HS-TCPPhoenix-PAIX Scalable-TCPPhoenix-PAIX Scalable-TCP #2

10 Gbits/s throughput from SC2003 to Chicago & Amsterdam

0

1

2

3

4

5

6

7

8

9

10

11/19/0315:59

11/19/0316:13

11/19/0316:27

11/19/0316:42

11/19/0316:56

11/19/0317:11

11/19/0317:25 Date & Time

Throughput

Gbits/s

Router traffic to Abilele

Phoenix-Chicago

Phoenix-Amsterdam

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Traffic QoS Classes on GÉANT Backbone

Normal Traffic

Normal Traffic +

Less Than Best Effort 2.0 Gbit/s

Normal Traffic +

Radio Astronomy Data 500 Mbit/s

Normal Traffic +

Radio Astronomy Data +

Less Than Best Effort 2.0 Gbit/s

Max Throughput on 2.5 G PoS

JIVE VLBI Network Meeting 28 Jan 2004R. Hughes-Jones Manchester

Some Measurements made during ER2002

No LBE

0

2

4

6

8

10

12

14

16

18

20

0 20 40 60 80 100 120 140 160 180 200Transfer number

No.

Out

of

ord

er

0

5000

10000

15000

20000

25000

No

. L

ost

num_badorder

num_lost

With 1.8Gbit LBE

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

0 20 40 60 80 100 120 140 160Transfer number

No.

Out

of

ord

er

0

5000

10000

15000

20000

25000

No

. L

ost

num_badorder

num_lost