1.4.1. RTI Routing Service¶

The following tests have been performed by executing the RTI Perftest C++98 benchmark application between two nodes and RTI Routing Service in a third node, all of them connected to a switch via Ethernet. The communication has been restricted to a single interface, and the transport has been set to UDPv4 for the three applications.

In this test, communication between the RTI Perftest Publisher and Subscriber is not direct, all traffic is routed with Routing Service. The Publisher application sends the samples, which are received and routed using Routing Service to the Subscriber application. If the sample needs to be returned to the Publisher application, the Subscriber sends it back, again through Routing Service. The latency measured is half Round Trip Time (RTT).

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

1.4.1.1. UDPv4¶

The graph below shows the one-way latency without load between a Publisher, a Routing Service, and a Subscriber running in three 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	44	18.8	38	9344	41	57	63	97	9344
64	45	20.7	38	9822	42	57	64	100	9822
128	45	17.2	38	8913	41	58	63	92	8913
512	47	20.0	40	9046	43	59	68	124	9046
1024	51	22.9	43	9217	47	62	78	99	9217
4096	60	5.6	55	120	57	66	83	114	120
8192	90	8.3	74	160	94	96	104	133	160
16384	195	42.3	95	330	192	247	295	322	330
32768	336	17.4	153	398	348	355	359	364	398
63000	376	6.0	218	446	376	379	384	400	446

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	45	8.9	36	133	40	60	63	70	133
64	45	9.0	36	121	41	60	63	70	121
128	46	8.2	36	91	43	60	62	69	91
512	46	357.6	38	100077	41	58	62	71	100077
1024	49	7.5	40	95	47	61	63	70	95
4096	58	1.8	55	125	58	61	63	75	125
8192	75	1.8	72	128	74	78	79	87	128
16384	189	30.2	99	346	186	214	336	342	346
32768	347	9.9	139	396	350	352	354	364	396
63000	371	4.7	206	396	371	373	375	383	396

Test Hardware

The following hardware was used to perform these tests:

Linux Nodes

Dell R340 Servers (13 Units)
Processor: Intel Xeon E-2278G (3.4-5GHz, 8c/16t, 16MB cache, 2 memory channels @2666MHz)
RAM: 4x 16GB 2666MHz DIMM (64GB RAM)
HD: 480GB SATA SSD
NIC 1: Intel 710 dual port 10Gbps SFP
OS: Ubuntu 20.04 -- gcc 9.3.0

Switch

Dell 2048 -- 10Gbps switch (10Gbps and 1Gbps interfaces)

The graph below shows the expected throughput behavior when performing a 1-1 communication between a Publisher, a Routing Service, and a Subscriber running in three Linux nodes in a 1Gbps network. The numbers are for best-effort as well as strict reliable reliability scenarios.

Note

By default, RTI Perftest enables batching when performing a Maximum Throughput test. The batching feature allows sending more than one data sample per RTPS packet, improving network performance for small data sizes. See the RTI Connext Core Libraries User’s Manual for more information on batching.

The batch maximum size is set by RTI Perftest to be 8192 bytes; after 8192 bytes, batching is not enabled.

Detailed Statistics

This table contains the raw numbers presented by RTI Perftest. These numbers are the exact output with no further processing.

Best Effort

Sample Size (Bytes)	Total Samples	Avg Samples/s	Avg Mbps	Lost Samples	Lost Samples (%)
32	5595904	268893	68.8	94404096	94.40
64	6168064	272913	139.7	93831936	93.83
128	7030528	275728	282.3	92969472	92.97
512	8221840	273150	1118.8	58039440	87.59
1024	7630800	253513	2276.8	28074584	78.63
4096	6206224	206185	6756.3	2771696	30.87
8192	4508374	150259	9847.4	0	0.00
16384	2263408	75438	9887.9	49	0.00
32768	1134034	37796	9908.3	21	0.00
63000	590258	19672	9915.2	7	0.00

Reliable

Sample Size (Bytes)	Total Samples	Avg Samples/s	Avg Mbps
32	4875265	161788	41.4
64	4939009	164220	84.1
128	4927617	163964	167.9
512	4859856	161722	662.4
1024	4695753	156282	1280.3
4096	4178251	139098	4558.0
8192	3660240	121826	7984.0
16384	2259808	75326	9873.1
32768	1133396	37776	9902.8
63000	590081	19668	9912.7

Test Hardware

The following hardware was used to perform these tests:

Linux Nodes

Dell R340 Servers (13 Units)
Processor: Intel Xeon E-2278G (3.4-5GHz, 8c/16t, 16MB cache, 2 memory channels @2666MHz)
RAM: 4x 16GB 2666MHz DIMM (64GB RAM)
HD: 480GB SATA SSD
NIC 1: Intel 710 dual port 10Gbps SFP
OS: Ubuntu 20.04 -- gcc 9.3.0

Switch

Dell 2048 -- 10Gbps switch (10Gbps and 1Gbps interfaces)

The graph below shows the one-way latency without load between a Publisher, a Routing Service, and a Subscriber running in three 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	47	5.7	41	107	44	57	63	89	107
64	47	30.8	41	9803	44	57	64	95	9803
128	48	19.5	42	8672	45	58	65	107	8672
512	50	11.9	43	3805	47	59	67	98	3805
1024	54	10.7	46	2671	50	62	92	114	2671
4096	65	5.6	59	144	64	70	90	118	144
8192	81	3.6	78	143	81	84	100	136	143
16384	194	61.0	99	351	194	276	317	341	351
32768	333	16.9	157	369	323	357	361	366	369
63000	341	5.8	219	421	340	346	356	406	421

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	47	7.9	39	126	42	60	63	120	126
64	46	8.4	39	127	41	59	69	118	127
128	52	312.6	35	100100	49	62	73	89	100100
512	50	7.0	41	173	47	61	63	69	173
1024	51	6.5	43	95	48	61	63	70	95
4096	62	1.4	58	101	62	63	66	74	101
8192	89	8.3	75	126	94	96	97	105	126
16384	194	42.9	97	344	183	249	321	327	344
32768	321	3.3	167	353	321	323	325	341	353
63000	377	5.5	210	391	377	380	382	389	391

Test Hardware

The following hardware was used to perform these tests:

Linux Nodes

Dell R340 Servers (13 Units)
Processor: Intel Xeon E-2278G (3.4-5GHz, 8c/16t, 16MB cache, 2 memory channels @2666MHz)
RAM: 4x 16GB 2666MHz DIMM (64GB RAM)
HD: 480GB SATA SSD
NIC 1: Intel 710 dual port 10Gbps SFP
OS: Ubuntu 20.04 -- gcc 9.3.0

Switch

Dell 2048 -- 10Gbps switch (10Gbps and 1Gbps interfaces)

The graph below shows the expected throughput behavior when performing a 1-1 communication between a Publisher, a Routing Service, and a Subscriber running in three Linux nodes in a 1Gbps network. The numbers are for best-effort as well as strict reliable reliability scenarios.

Note

By default, RTI Perftest enables batching when performing a Maximum Throughput test. The batching feature allows sending more than one data sample per RTPS packet, improving network performance for small data sizes. See the RTI Connext Core Libraries User’s Manual for more information on batching.

The batch maximum size is set by RTI Perftest to be 8192 bytes; after 8192 bytes, batching is not enabled.

Detailed Statistics

This table contains the raw numbers presented by RTI Perftest. These numbers are the exact output with no further processing.

Best Effort

Sample Size (Bytes)	Total Samples	Avg Samples/s	Avg Mbps	Lost Samples	Lost Samples (%)
32	6747392	224688	57.5	47125248	87.48
64	6920832	230453	118.0	45027435	86.68
128	6888649	229352	234.9	41392704	85.73
512	6385376	212139	968.9	30000656	82.45
1024	5911452	196909	1813.1	20732184	77.81
4096	5893528	196231	6430.1	3032924	33.98
8192	4495430	149819	9818.6	0	0.00
16384	2260179	75329	9873.6	31	0.00
32768	1133222	37769	9901.1	14	0.00
63000	590034	19665	9911.5	9	0.00

Reliable

Sample Size (Bytes)	Total Samples	Avg Samples/s	Avg Mbps
32	4529665	150501	38.5
64	4410497	146732	75.1
128	4383885	145911	149.4
512	4456353	148344	607.6
1024	4325905	143964	1179.4
4096	3957422	131696	4315.4
8192	3491797	116225	7617.0
16384	2256941	75230	9860.6
32768	1132571	37748	9895.6
63000	589861	19660	9908.9

Test Hardware

The following hardware was used to perform these tests:

Linux Nodes

Dell R340 Servers (13 Units)
Processor: Intel Xeon E-2278G (3.4-5GHz, 8c/16t, 16MB cache, 2 memory channels @2666MHz)
RAM: 4x 16GB 2666MHz DIMM (64GB RAM)
HD: 480GB SATA SSD
NIC 1: Intel 710 dual port 10Gbps SFP
OS: Ubuntu 20.04 -- gcc 9.3.0

Switch

Dell 2048 -- 10Gbps switch (10Gbps and 1Gbps interfaces)