* [dts] [DTS][PATCH V3 2/3]: test_plans/nic_single_core_perf_test_plan.rst: add test plan for nic_single_core_perf
@ 2018-01-30 10:43 wang fei
2018-01-31 7:29 ` Liu, Yong
0 siblings, 1 reply; 2+ messages in thread
From: wang fei @ 2018-01-30 10:43 UTC (permalink / raw)
To: dts; +Cc: yulong.pei, wang fei
For next branch: add the test plan for the test suite nic_single_core_perf
Signed-off-by: wang fei <feix.y.wang@intel.com>
---
test_plans/nic_single_core_perf_test_plan.rst | 116 ++++++++++++++++++++++++++
1 file changed, 116 insertions(+)
create mode 100644 test_plans/nic_single_core_perf_test_plan.rst
diff --git a/test_plans/nic_single_core_perf_test_plan.rst b/test_plans/nic_single_core_perf_test_plan.rst
new file mode 100644
index 0000000..ec67ad4
--- /dev/null
+++ b/test_plans/nic_single_core_perf_test_plan.rst
@@ -0,0 +1,116 @@
+.. Copyright (c) <2016>, Intel Corporation
+ All rights reserved.
+
+ Redistribution and use in source and binary forms, with or without
+ modification, are permitted provided that the following conditions
+ are met:
+
+ - Redistributions of source code must retain the above copyright
+ notice, this list of conditions and the following disclaimer.
+
+ - Redistributions in binary form must reproduce the above copyright
+ notice, this list of conditions and the following disclaimer in
+ the documentation and/or other materials provided with the
+ distribution.
+
+ - Neither the name of Intel Corporation nor the names of its
+ contributors may be used to endorse or promote products derived
+ from this software without specific prior written permission.
+
+ THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
+ FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
+ COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
+ INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+ SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
+ HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
+ STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
+ ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
+ OF THE POSSIBILITY OF SUCH DAMAGE.
+
+==========================================================================
+Benchmark the performance of single core forwarding with FVL25G/40G/NNT10G
+==========================================================================
+
+Prerequisites
+=============
+
+1. Hardware:
+ 1) nic_single_core_perf test for FVL25G: two dual port FVL25G nics , all installed on the same socket, pick one port per nic
+ 2) nic_single_core_perf test for FVL40G: two dual port FVL40G nics , all installed on the same socket, pick one port per nic
+ 3) nic_single_core_perf test for NNT10G : four 82599 nics, all installed on the same socket, pick one port per nic
+
+2. Software:
+ dpdk: git clone http://dpdk.org/git/dpdk
+ scapy: http://www.secdev.org/projects/scapy/
+ dts (next branch): git clone http://dpdk.org/git/tools/dts, then "git checkout next"
+ Trex source code: http://trex-tgn.cisco.com/trex/release/v2.26.tar.gz
+ python-prettytable(uesed to output the test result matrix as a separate .txt file):
+ apt install python-prettytable (for ubuntu os)
+ or dnf install python-prettytable (for fedora os).
+
+3. Connect all the selected nic ports to traffic generator(IXIA,TREX,PKTGEN) ports(TG ports).
+ 2 TG 25g ports for FVL25G ports
+ 2 TG 40g ports for FVL40G ports
+ 4 TG 10g ports for 4 NNT10G ports
+
+Test Case : nic_single_core_perf
+==================================
+1) Bind tested ports to igb_uio
+
+2) Start testpmd::
+
+ ./x86_64-native-linuxapp-gcc/app/testpmd -c 0x6 -n 4 -- -i --portmask=0xf --txd=128 --rxd=128
+ testpmd> start
+
+3) Configure traffic generator to send traffic
+ configure one below stream for each TG port
+ dst mac: peer nic port mac
+ src ip : random
+ dst ip : random
+ packet length : 64 byte
+
+4) check throughput and compare it with the expected value.
+
+5) for NNT10G, repeat above step 1-4 for txd=rxd=512,2048 separately.
+ for FVL25G and FVL40G nic, just test txd=rxd=512,2048 following above steps 1-4.
+
+6) print out below test results table for NNT10G, FVL25G, FVL40G separately
+ FVL40G:
++------------+---------+------------+------------+---------------------+
+| Frame Size | txd/rxd | Throughput | Rate | Expected throughput |
++============+=========+============+============+=====================+
+| 64 | 512 | xxxxxx | xxx % | xxx Mpps |
++------------+---------+------------+------------+---------------------+
+| 64 | 2048 | xxxxxx | xxx % | xxx Mpps |
++------------+---------+------------+------------+---------------------+
+
+ FVL25G:
++------------+---------+-------------+---------+---------------------+
+| Frame Size | TXD/RXD | Throughput | Rate | Expected Throughput |
++------------+---------+-------------+---------+---------------------+
+| 64 | 512 | xxxxxx Mpps | xxx % | xxx Mpps |
+| 64 | 2048 | xxxxxx Mpps | xxx % | xxx Mpps |
++------------+---------+-------------+---------+---------------------+
+
+ for NNT10G:
++------------+---------+-------------+---------+---------------------+
+| Frame Size | TXD/RXD | Throughput | Rate | Expected Throughput |
++------------+---------+-------------+---------+---------------------+
+| 64 | 128 | xxxxxx Mpps | xxx % | xxx Mpps |
+| 64 | 512 | xxxxxx Mpps | xxx % | xxx Mpps |
+| 64 | 2048 | xxxxxx Mpps | xxx % | xxx Mpps |
++------------+---------+-------------+---------+---------------------+
+
+Note : The values for the Expected throughput may vary due to different platform and OS, and traffic generator,
+ please correct thest values accordingly. (the above FVL 25G and NNT10G expected throughput values
+ were got from the combination of Purly,Ubuntu 16.04, and traffic generator IXIA, and the FVL40G
+ expected throughput value was got from the combination of Broadwell,Ubuntu 16.04, and traffic
+ generator IXIA)
+
+Case assert failure if the actual throughputs lower exceed 1 Mpps than the expected throughput.
+
+
+
--
2.7.4
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [dts] [DTS][PATCH V3 2/3]: test_plans/nic_single_core_perf_test_plan.rst: add test plan for nic_single_core_perf
2018-01-30 10:43 [dts] [DTS][PATCH V3 2/3]: test_plans/nic_single_core_perf_test_plan.rst: add test plan for nic_single_core_perf wang fei
@ 2018-01-31 7:29 ` Liu, Yong
0 siblings, 0 replies; 2+ messages in thread
From: Liu, Yong @ 2018-01-31 7:29 UTC (permalink / raw)
To: Wang, FeiX Y, dts; +Cc: Pei, Yulong, Wang, FeiX Y
Fei,
Some comments are inline. Please try to keep each line in 79 characters.
> -----Original Message-----
> From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of wang fei
> Sent: Tuesday, January 30, 2018 6:43 PM
> To: dts@dpdk.org
> Cc: Pei, Yulong <yulong.pei@intel.com>; Wang, FeiX Y
> <feix.y.wang@intel.com>
> Subject: [dts] [DTS][PATCH V3 2/3]:
> test_plans/nic_single_core_perf_test_plan.rst: add test plan for
> nic_single_core_perf
>
> For next branch: add the test plan for the test suite nic_single_core_perf
>
> Signed-off-by: wang fei <feix.y.wang@intel.com>
> ---
> test_plans/nic_single_core_perf_test_plan.rst | 116
> ++++++++++++++++++++++++++
> 1 file changed, 116 insertions(+)
> create mode 100644 test_plans/nic_single_core_perf_test_plan.rst
>
> diff --git a/test_plans/nic_single_core_perf_test_plan.rst
> b/test_plans/nic_single_core_perf_test_plan.rst
> new file mode 100644
> index 0000000..ec67ad4
> --- /dev/null
> +++ b/test_plans/nic_single_core_perf_test_plan.rst
> @@ -0,0 +1,116 @@
> +.. Copyright (c) <2016>, Intel Corporation
> + All rights reserved.
> +
> + Redistribution and use in source and binary forms, with or without
> + modification, are permitted provided that the following conditions
> + are met:
> +
> + - Redistributions of source code must retain the above copyright
> + notice, this list of conditions and the following disclaimer.
> +
> + - Redistributions in binary form must reproduce the above copyright
> + notice, this list of conditions and the following disclaimer in
> + the documentation and/or other materials provided with the
> + distribution.
> +
> + - Neither the name of Intel Corporation nor the names of its
> + contributors may be used to endorse or promote products derived
> + from this software without specific prior written permission.
> +
> + THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> + "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> + LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
> + FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> + COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> + INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
> + (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
> + SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
> + HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT,
> + STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
> + ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED
> + OF THE POSSIBILITY OF SUCH DAMAGE.
> +
> +=========================================================================
> =
> +Benchmark the performance of single core forwarding with
> FVL25G/40G/NNT10G
> +=========================================================================
> =
> +
> +Prerequisites
> +=============
> +
> +1. Hardware:
> + 1) nic_single_core_perf test for FVL25G: two dual port FVL25G nics ,
> all installed on the same socket, pick one port per nic
I guess that all NIC should installed on socket 0. So the statement will be like:
FVL25: Two dual ports FVL25 NICs, both installed on socket 0.
> + 2) nic_single_core_perf test for FVL40G: two dual port FVL40G nics ,
> all installed on the same socket, pick one port per nic
> + 3) nic_single_core_perf test for NNT10G : four 82599 nics, all
> installed on the same socket, pick one port per nic
> +
> +2. Software:
> + dpdk: git clone http://dpdk.org/git/dpdk
> + scapy: http://www.secdev.org/projects/scapy/
> + dts (next branch): git clone http://dpdk.org/git/tools/dts, then "git
> checkout next"
> + Trex source code: http://trex-tgn.cisco.com/trex/release/v2.26.tar.gz
> + python-prettytable(uesed to output the test result matrix as a
> separate .txt file):
> + apt install python-prettytable (for ubuntu os)
> + or dnf install python-prettytable (for fedora os).
> +
> +3. Connect all the selected nic ports to traffic
> generator(IXIA,TREX,PKTGEN) ports(TG ports).
> + 2 TG 25g ports for FVL25G ports
> + 2 TG 40g ports for FVL40G ports
> + 4 TG 10g ports for 4 NNT10G ports
> +
> +Test Case : nic_single_core_perf
> +==================================
Please use case name "Single Core Performance Measurement".
> +1) Bind tested ports to igb_uio
> +
> +2) Start testpmd::
> +
> + ./x86_64-native-linuxapp-gcc/app/testpmd -c 0x6 -n 4 -- -i --
> portmask=0xf --txd=128 --rxd=128
> + testpmd> start
> +
> +3) Configure traffic generator to send traffic
> + configure one below stream for each TG port
> + dst mac: peer nic port mac
> + src ip : random
> + dst ip : random
> + packet length : 64 byte
> +
> +4) check throughput and compare it with the expected value.
> +
> +5) for NNT10G, repeat above step 1-4 for txd=rxd=512,2048 separately.
> + for FVL25G and FVL40G nic, just test txd=rxd=512,2048 following above
> steps 1-4.
> +
> +6) print out below test results table for NNT10G, FVL25G, FVL40G
> separately
Simplify the description like:
Result tables for different NIC:
> + FVL40G:
> ++------------+---------+------------+------------+---------------------+
> +| Frame Size | txd/rxd | Throughput | Rate | Expected throughput |
> ++============+=========+============+============+=====================+
> +| 64 | 512 | xxxxxx | xxx % | xxx Mpps |
> ++------------+---------+------------+------------+---------------------+
> +| 64 | 2048 | xxxxxx | xxx % | xxx Mpps |
> ++------------+---------+------------+------------+---------------------+
> +
> + FVL25G:
> ++------------+---------+-------------+---------+---------------------+
> +| Frame Size | TXD/RXD | Throughput | Rate | Expected Throughput |
> ++------------+---------+-------------+---------+---------------------+
> +| 64 | 512 | xxxxxx Mpps | xxx % | xxx Mpps |
> +| 64 | 2048 | xxxxxx Mpps | xxx % | xxx Mpps |
> ++------------+---------+-------------+---------+---------------------+
> +
> + for NNT10G:
> ++------------+---------+-------------+---------+---------------------+
> +| Frame Size | TXD/RXD | Throughput | Rate | Expected Throughput |
> ++------------+---------+-------------+---------+---------------------+
> +| 64 | 128 | xxxxxx Mpps | xxx % | xxx Mpps |
> +| 64 | 512 | xxxxxx Mpps | xxx % | xxx Mpps |
> +| 64 | 2048 | xxxxxx Mpps | xxx % | xxx Mpps |
> ++------------+---------+-------------+---------+---------------------+
> +
> +Note : The values for the Expected throughput may vary due to different
> platform and OS, and traffic generator,
Please take care the grammar. "Expected" should be "expected".
> + please correct thest values accordingly. (the above FVL 25G and
> NNT10G expected throughput values
"thest" should "threshold".
> + were got from the combination of Purly,Ubuntu 16.04, and traffic
> generator IXIA, and the FVL40G
> + expected throughput value was got from the combination of
> Broadwell,Ubuntu 16.04, and traffic
> + generator IXIA)
Why the criteria performance stats are from different settings?
Open-lab setting are different for those NICs?
> +
> +Case assert failure if the actual throughputs lower exceed 1 Mpps than
> the expected throughput.
> +
Please use below sentence.
Case will raise failure if actual throughputs have more than 1Mpps gap from expected ones.
> +
> +
> --
> 2.7.4
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2018-01-31 7:29 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2018-01-30 10:43 [dts] [DTS][PATCH V3 2/3]: test_plans/nic_single_core_perf_test_plan.rst: add test plan for nic_single_core_perf wang fei
2018-01-31 7:29 ` Liu, Yong
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).