1.2.3. Java API

The following tests have been performed by executing an RTI Perftest Java Publisher and Subscriber between two nodes, connected to a switch via Ethernet. The communication has been restricted to a single interface and the transport has been set to UDPv4.

Find information about the hardware, network, and command-line parameters after each of the tests.

1.2.3.1. Unkeyed, UDPv4 10Gbps Network, Java

The graph below shows the one-way latency without load between a Publisher and a Subscriber running in two Linux nodes in a 10Gbps network. The numbers are for best-effort as well as strict reliable reliability scenarios.

Note

We use the median (50th percentile) instead of the average in order to get a more stable measurement that does not account for spurious outliers. We also calculate the average value and other percentile values, which can be seen in the Detailed Statistics section below.

Detailed Statistics

The following tables contain the raw numbers presented by RTI Perftest. These numbers are the exact output with no further processing.

  • Best Effort

Sample Size (Bytes)

Avg (μs)

Std (μs)

Min (μs)

Max (μs)

50% (μs)

90% (μs)

99% (μs)

99.99% (μs)

99.9999% (μs)

32

27

11.0

26

6418

26

27

33

26

26

64

26

10.6

25

6344

26

27

33

25

25

128

27

10.8

26

6385

26

28

34

26

26

256

28

10.9

27

6247

27

28

35

27

27

512

28

11.0

27

6368

27

29

35

27

27

1024

30

11.2

29

6411

30

30

37

29

29

8192

51

14.1

49

6235

49

51

58

49

49

63000

117

20.7

115

6453

115

118

126

115

115

  • Reliable

Sample Size (Bytes)

Avg (μs)

Std (μs)

Min (μs)

Max (μs)

50% (μs)

90% (μs)

99% (μs)

99.99% (μs)

99.9999% (μs)

32

27

19.4

26

13062

27

28

35

26

26

64

29

11.2

28

6230

28

30

37

28

28

128

29

11.1

27

6297

27

30

37

27

27

256

30

11.4

28

6217

28

31

41

28

28

512

31

11.7

29

6366

28

32

42

29

29

1024

32

11.5

30

6226

30

32

42

30

30

8192

52

14.8

50

6370

50

53

59

50

50

63000

121

21.2

117

6416

117

126

133

117

117

100000

175

23.7

160

6530

160

187

195

160

160

500000

656

59.1

547

7191

624

698

727

547

547

1048576

1289

94.7

1110

7687

1270

1403

1494

1110

1110

1548576

2012

162.5

1678

8477

1934

2188

2304

1678

1678

4194304

6044

2124.3

5411

40574

5549

5934

10539

5411

5411

10485760

15889

5608.0

14490

50095

14802

15397

49792

14490

14490


Perftest Scripts

To produce these tests, we executed RTI Perftest for C++98. The exact commands used can be found here:

Publisher Side

 1sudo /set_lat_mode.sh
 2
 3echo EXECUTABLE IS $1
 4export executable=$1
 5
 6echo OUTPUT PATH IS $2
 7export output_folder=$2
 8
 9export exec_time=30
10export nic=172.16.0.1
11export pub_string="-pub \
12        -transport UDPv4 \
13        -nic $nic \
14        -noPrint \
15        -noOutputHeaders \
16        -exec $exec_time \
17        -latencyTest"
18
19mkdir -p $output_folder
20
21echo ">> UNKEYED BE"
22export my_file=$output_folder/lat_udpv4_pub_unkeyed_be.csv
23touch $my_file
24for DATALEN in 32 64 128 256 512 1024 8192 63000; do
25    export command="taskset -c 0 \
26    $executable -best -datalen $DATALEN $pub_string"
27    echo $command
28    $command >> $my_file;
29    sleep 3;
30done
31sleep 5;
32
33echo ">> UNKEYED REL"
34export my_file=$output_folder/lat_udpv4_pub_unkeyed_rel.csv
35touch $my_file
36for DATALEN in 32 64 128 256 512 1024 8192 63000 100000 500000 1048576 1548576 4194304 10485760; do
37    export command="taskset -c 0 \
38    $executable -datalen $DATALEN $pub_string"
39    echo $command
40    $command >> $my_file;
41    sleep 3;
42done
43sleep 5;
44
45echo ">> KEYED BE"
46export my_file=$output_folder/lat_udpv4_pub_keyed_be.csv
47touch $my_file
48for DATALEN in 32 64 128 256 512 1024 8192 63000; do
49    export command="taskset -c 0 \
50    $executable -best -keyed -instances 100000 -datalen $DATALEN $pub_string"
51    echo $command
52    $command >> $my_file;
53    sleep 3;
54done
55sleep 5;
56
57echo ">> KEYED REL"
58export my_file=$output_folder/lat_udpv4_pub_keyed_rel.csv
59touch $my_file
60for DATALEN in 32 64 128 256 512 1024 8192 63000; do
61    export command="taskset -c 0 \
62    $executable -keyed -instances 100000 -datalen $DATALEN $pub_string"
63    echo $command
64    $command >> $my_file;
65    sleep 3;
66done

Subscriber Side

 1sudo /set_lat_mode.sh
 2
 3echo EXECUTABLE IS $1
 4export executable=$1
 5
 6echo OUTPUT PATH IS $2
 7export output_folder=$2
 8
 9export nic=172.16.0.2
10export sub_string="-sub \
11        -transport UDPv4 \
12        -nic $nic \
13        -noPrint \
14        -noOutputHeaders"
15
16mkdir -p $output_folder
17
18echo ">> UNKEYED BE"
19export my_file=$output_folder/lat_udpv4_sub_unkeyed_be.csv
20touch $my_file
21for DATALEN in 32 64 128 256 512 1024 8192 63000; do
22    export command="taskset -c 0 \
23    $executable -best $sub_string -datalen $DATALEN"
24    echo $command
25    $command >> $my_file;
26    sleep 10;
27done
28sleep 5;
29
30echo ">> UNKEYED REL"
31export my_file=$output_folder/lat_udpv4_sub_unkeyed_rel.csv
32touch $my_file
33for DATALEN in 32 64 128 256 512 1024 8192 63000 100000 500000 1048576 1548576 4194304 10485760; do
34    export command="taskset -c 0 \
35    $executable $sub_string -datalen $DATALEN"
36    echo $command
37    $command >> $my_file;
38    sleep 10;
39done
40sleep 5;
41
42echo ">> KEYED BE"
43export my_file=$output_folder/lat_udpv4_sub_keyed_be.csv
44touch $my_file
45for DATALEN in 32 64 128 256 512 1024 8192 63000; do
46    export command="taskset -c 0 \
47    $executable -best -keyed -instances 100000 $sub_string -datalen $DATALEN"
48    echo $command
49    $command >> $my_file;
50    sleep 10;
51done
52sleep 5;
53
54echo ">> KEYED REL"
55export my_file=$output_folder/lat_udpv4_sub_keyed_rel.csv
56touch $my_file
57for DATALEN in 32 64 128 256 512 1024 8192 63000; do
58    export command="taskset -c 0 \
59    $executable -keyed -instances 100000 $sub_string -datalen $DATALEN"
60    echo $command
61    $command >> $my_file;
62    sleep 10;
63done

Test Hardware

The following hardware was used to perform these tests:

Linux Nodes

Processor: Intel® Xeon® E-2186G 3.8GHz, 12M cache, 6C/12T, turbo (95W)
RAM: 16GB 2666MT/s DDR4 ECC UDIMM
NIC 1: Intel X550 Dual Port 10GbE BASE-T Adapter, PCIe Full Height
NIC 2: Intel Ethernet I350 Dual Port 1GbE BASE-T Adapter, PCIe Low Profile
OS: Ubuntu 18.04 -- gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0

Switch

Dell Networking S4048T-ON, 48x 10GBASE-T and 6x 40GbE QSFP+ ports, IO to PSU air, 2x AC PSU, OS9

1.2.3.2. Keyed, UDPv4 10Gbps Network, Java

The graph below shows the one-way latency without load between a Publisher and a Subscriber running in two Linux nodes in a 10Gbps network. The numbers are for best-effort as well as strict reliable reliability scenarios.

Note

We use the median (50th percentile) instead of the average in order to get a more stable measurement that does not account for spurious outliers. We also calculate the average value and other percentile values, which can be seen in the Detailed Statistics section below.

Detailed Statistics

The following tables contain the raw numbers presented by RTI Perftest. These numbers are the exact output with no further processing.

  • Best Effort

Sample Size (Bytes)

Avg (μs)

Std (μs)

Min (μs)

Max (μs)

50% (μs)

90% (μs)

99% (μs)

99.99% (μs)

99.9999% (μs)

32

26

18.8

25

13070

26

26

33

25

25

64

26

19.1

25

13183

26

26

33

25

25

128

27

19.2

25

13167

26

27

32

25

25

256

27

19.3

26

13210

27

27

33

26

26

512

28

19.6

27

13184

28

28

35

27

27

1024

29

20.0

28

13204

29

29

34

28

28

8192

50

25.4

48

13181

49

50

57

48

48

63000

117

38.9

115

13312

116

118

125

115

115

  • Reliable

Sample Size (Bytes)

Avg (μs)

Std (μs)

Min (μs)

Max (μs)

50% (μs)

90% (μs)

99% (μs)

99.99% (μs)

99.9999% (μs)

32

32

23.8

31

9342

32

33

36

31

31

64

32

24.6

31

9817

32

33

36

31

31

128

33

23.9

31

9484

32

33

36

31

31

256

34

28.4

32

13076

33

34

37

32

32

512

34

27.7

32

12385

33

34

37

32

32

1024

36

24.4

34

9447

35

36

39

34

34

8192

56

34.9

54

12481

56

57

59

54

54

63000

125

54.2

121

13578

124

126

131

121

121


Perftest Scripts

To produce these tests, we executed RTI Perftest for C++98. The exact commands used can be found here:

Publisher Side

Subscriber Side


Test Hardware

The following hardware was used to perform these tests:

Linux Nodes

Processor: Intel® Xeon® E-2186G 3.8GHz, 12M cache, 6C/12T, turbo (95W)
RAM: 16GB 2666MT/s DDR4 ECC UDIMM
NIC 1: Intel X550 Dual Port 10GbE BASE-T Adapter, PCIe Full Height
NIC 2: Intel Ethernet I350 Dual Port 1GbE BASE-T Adapter, PCIe Low Profile
OS: Ubuntu 18.04 -- gcc (Ubuntu 7.5.0-3ubuntu1~18.04) 7.5.0

Switch

Dell Networking S4048T-ON, 48x 10GBASE-T and 6x 40GbE QSFP+ ports, IO to PSU air, 2x AC PSU, OS9