1 .. This work is licensed under a Creative Commons Attribution 4.0 International License.
2 .. http://creativecommons.org/licenses/by/4.0
3 .. (c) OPNFV, Intel Corporation, AT&T and others.
5 vSwitchPerf test suites userguide
6 ---------------------------------
11 VSPERF requires a traffic generators to run tests, automated traffic gen
12 support in VSPERF includes:
14 - IXIA traffic generator (IxNetwork hardware) and a machine that runs the IXIA
16 - Spirent traffic generator (TestCenter hardware chassis or TestCenter virtual
17 in a VM) and a VM to run the Spirent Virtual Deployment Service image,
18 formerly known as "Spirent LabServer".
19 - Xena Network traffic generator (Xena hardware chassis) that houses the Xena
20 Traffic generator modules.
21 - Moongen software traffic generator. Requires a separate machine running
22 moongen to execute packet generation.
23 - T-Rex software traffic generator. Requires a separate machine running T-Rex
24 Server to execute packet generation.
26 If you want to use another traffic generator, please select the :ref:`trafficgen-dummy`
32 To see the supported Operating Systems, vSwitches and system requirements,
33 please follow the `installation instructions <vsperf-installation>`.
35 Traffic Generator Setup
36 ^^^^^^^^^^^^^^^^^^^^^^^
38 Follow the `Traffic generator instructions <trafficgen-installation>` to
39 install and configure a suitable traffic generator.
41 Cloning and building src dependencies
42 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
44 In order to run VSPERF, you will need to download DPDK and OVS. You can
45 do this manually and build them in a preferred location, OR you could
46 use vswitchperf/src. The vswitchperf/src directory contains makefiles
47 that will allow you to clone and build the libraries that VSPERF depends
48 on, such as DPDK and OVS. To clone and build simply:
50 .. code-block:: console
55 VSPERF can be used with stock OVS (without DPDK support). When build
56 is finished, the libraries are stored in src_vanilla directory.
58 The 'make' builds all options in src:
61 * OVS with vhost_user as the guest access method (with DPDK support)
63 The vhost_user build will reside in src/ovs/
64 The Vanilla OVS build will reside in vswitchperf/src_vanilla
66 To delete a src subdirectory and its contents to allow you to re-clone simply
69 .. code-block:: console
73 Configure the ``./conf/10_custom.conf`` file
74 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
76 The ``10_custom.conf`` file is the configuration file that overrides
77 default configurations in all the other configuration files in ``./conf``
78 The supplied ``10_custom.conf`` file **MUST** be modified, as it contains
79 configuration items for which there are no reasonable default values.
81 The configuration items that can be added is not limited to the initial
82 contents. Any configuration item mentioned in any .conf file in
83 ``./conf`` directory can be added and that item will be overridden by
84 the custom configuration value.
86 Further details about configuration files evaluation and special behaviour
87 of options with ``GUEST_`` prefix could be found at :ref:`design document
88 <design-configuration>`.
90 Using a custom settings file
91 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
93 If your ``10_custom.conf`` doesn't reside in the ``./conf`` directory
94 of if you want to use an alternative configuration file, the file can
95 be passed to ``vsperf`` via the ``--conf-file`` argument.
97 .. code-block:: console
99 $ ./vsperf --conf-file <path_to_custom_conf> ...
101 Note that configuration passed in via the environment (``--load-env``)
102 or via another command line argument will override both the default and
103 your custom configuration files. This "priority hierarchy" can be
104 described like so (1 = max priority):
106 1. Testcase definition section ``Parameters``
107 2. Command line arguments
108 3. Environment variables
109 4. Configuration file(s)
111 Further details about configuration files evaluation and special behaviour
112 of options with ``GUEST_`` prefix could be found at :ref:`design document
113 <design-configuration>`.
115 .. _overriding-parameters-documentation:
117 Referencing parameter values
118 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
120 It is possible to use a special macro ``#PARAM()`` to refer to the value of
121 another configuration parameter. This reference is evaluated during
122 access of the parameter value (by ``settings.getValue()`` call), so it
123 can refer to parameters created during VSPERF runtime, e.g. NICS dictionary.
124 It can be used to reflect DUT HW details in the testcase definition.
136 # set destination MAC to the MAC of the first
137 # interface from WHITELIST_NICS list
138 'dstmac' : '#PARAM(NICS[0]["mac"])',
143 Overriding values defined in configuration files
144 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
146 The configuration items can be overridden by command line argument
147 ``--test-params``. In this case, the configuration items and
148 their values should be passed in form of ``item=value`` and separated
155 $ ./vsperf --test-params "TRAFFICGEN_DURATION=10;TRAFFICGEN_PKT_SIZES=(128,);" \
156 "GUEST_LOOPBACK=['testpmd','l2fwd']" pvvp_tput
158 The ``--test-params`` command line argument can also be used to override default
159 configuration values for multiple tests. Providing a list of parameters will apply each
160 element of the list to the test with the same index. If more tests are run than
161 parameters provided the last element of the list will repeat.
165 $ ./vsperf --test-params "['TRAFFICGEN_DURATION=10;TRAFFICGEN_PKT_SIZES=(128,)',"
166 "'TRAFFICGEN_DURATION=10;TRAFFICGEN_PKT_SIZES=(64,)']" \
169 The second option is to override configuration items by ``Parameters`` section
170 of the test case definition. The configuration items can be added into ``Parameters``
171 dictionary with their new values. These values will override values defined in
172 configuration files or specified by ``--test-params`` command line argument.
178 "Parameters" : {'TRAFFICGEN_PKT_SIZES' : (128,),
179 'TRAFFICGEN_DURATION' : 10,
180 'GUEST_LOOPBACK' : ['testpmd','l2fwd'],
183 **NOTE:** In both cases, configuration item names and their values must be specified
184 in the same form as they are defined inside configuration files. Parameter names
185 must be specified in uppercase and data types of original and new value must match.
186 Python syntax rules related to data types and structures must be followed.
187 For example, parameter ``TRAFFICGEN_PKT_SIZES`` above is defined as a tuple
188 with a single value ``128``. In this case trailing comma is mandatory, otherwise
189 value can be wrongly interpreted as a number instead of a tuple and vsperf
190 execution would fail. Please check configuration files for default values and their
191 types and use them as a basis for any customized values. In case of any doubt, please
192 check official python documentation related to data structures like tuples, lists
195 **NOTE:** Vsperf execution will terminate with runtime error in case, that unknown
196 parameter name is passed via ``--test-params`` CLI argument or defined in ``Parameters``
197 section of test case definition. It is also forbidden to redefine a value of
198 ``TEST_PARAMS`` configuration item via CLI or ``Parameters`` section.
203 VSPERF uses a VM image called vloop_vnf for looping traffic in the deployment
204 scenarios involving VMs. The image can be downloaded from
205 `<http://artifacts.opnfv.org/>`__.
207 Please see the installation instructions for information on :ref:`vloop-vnf`
215 A Kernel Module that provides OSI Layer 2 Ipv4 termination or forwarding with
216 support for Destination Network Address Translation (DNAT) for both the MAC and
217 IP addresses. l2fwd can be found in <vswitchperf_dir>/src/l2fwd
219 Additional Tools Setup
220 ^^^^^^^^^^^^^^^^^^^^^^
222 Follow the `Additional tools instructions <additional-tools-configuration>` to
223 install and configure additional tools such as collectors and loadgens.
228 All examples inside these docs assume, that user is inside the VSPERF
229 directory. VSPERF can be executed from any directory.
231 Before running any tests make sure you have root permissions by adding
232 the following line to /etc/sudoers:
234 .. code-block:: console
236 username ALL=(ALL) NOPASSWD: ALL
238 username in the example above should be replaced with a real username.
240 To list the available tests:
242 .. code-block:: console
246 To run a single test:
248 .. code-block:: console
252 Where $TESTNAME is the name of the vsperf test you would like to run.
254 To run a test multiple times, repeat it:
256 .. code-block:: console
258 $ ./vsperf $TESTNAME $TESTNAME $TESTNAME
260 To run a group of tests, for example all tests with a name containing
263 .. code-block:: console
265 $ ./vsperf --conf-file=<path_to_custom_conf>/10_custom.conf --tests="RFC2544"
269 .. code-block:: console
271 $ ./vsperf --conf-file=<path_to_custom_conf>/10_custom.conf
273 Some tests allow for configurable parameters, including test duration
274 (in seconds) as well as packet sizes (in bytes).
278 $ ./vsperf --conf-file user_settings.py \
279 --tests RFC2544Tput \
280 --test-params "TRAFFICGEN_DURATION=10;TRAFFICGEN_PKT_SIZES=(128,)"
282 To specify configurable parameters for multiple tests, use a list of
283 parameters. One element for each test.
287 $ ./vsperf --conf-file user_settings.py \
288 --test-params "['TRAFFICGEN_DURATION=10;TRAFFICGEN_PKT_SIZES=(128,)',"\
289 "'TRAFFICGEN_DURATION=10;TRAFFICGEN_PKT_SIZES=(64,)']" \
290 phy2phy_cont phy2phy_cont
292 If the ``CUMULATIVE_PARAMS`` setting is set to True and there are different parameters
293 provided for each test using ``--test-params``, each test will take the parameters of
294 the previous test before appyling it's own.
295 With ``CUMULATIVE_PARAMS`` set to True the following command will be equivalent to the
300 $ ./vsperf --conf-file user_settings.py \
301 --test-params "['TRAFFICGEN_DURATION=10;TRAFFICGEN_PKT_SIZES=(128,)',"\
302 "'TRAFFICGEN_PKT_SIZES=(64,)']" \
303 phy2phy_cont phy2phy_cont
306 For all available options, check out the help dialog:
308 .. code-block:: console
312 Executing Vanilla OVS tests
313 ^^^^^^^^^^^^^^^^^^^^^^^^^^^
315 1. If needed, recompile src for all OVS variants
317 .. code-block:: console
323 2. Update your ``10_custom.conf`` file to use Vanilla OVS:
325 .. code-block:: python
327 VSWITCH = 'OvsVanilla'
331 .. code-block:: console
333 $ ./vsperf --conf-file=<path_to_custom_conf>
335 Please note if you don't want to configure Vanilla OVS through the
336 configuration file, you can pass it as a CLI argument.
338 .. code-block:: console
340 $ ./vsperf --vswitch OvsVanilla
343 Executing tests with VMs
344 ^^^^^^^^^^^^^^^^^^^^^^^^
346 To run tests using vhost-user as guest access method:
348 1. Set VSWITCH and VNF of your settings file to:
350 .. code-block:: python
352 VSWITCH = 'OvsDpdkVhost'
353 VNF = 'QemuDpdkVhost'
355 2. If needed, recompile src for all OVS variants
357 .. code-block:: console
365 .. code-block:: console
367 $ ./vsperf --conf-file=<path_to_custom_conf>/10_custom.conf
369 **NOTE:** By default vSwitch is acting as a server for dpdk vhost-user sockets.
370 In case, that QEMU should be a server for vhost-user sockets, then parameter
371 ``VSWITCH_VHOSTUSER_SERVER_MODE`` should be set to ``False``.
373 Executing tests with VMs using Vanilla OVS
374 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
376 To run tests using Vanilla OVS:
378 1. Set the following variables:
380 .. code-block:: python
382 VSWITCH = 'OvsVanilla'
383 VNF = 'QemuVirtioNet'
385 VANILLA_TGEN_PORT1_IP = n.n.n.n
386 VANILLA_TGEN_PORT1_MAC = nn:nn:nn:nn:nn:nn
388 VANILLA_TGEN_PORT2_IP = n.n.n.n
389 VANILLA_TGEN_PORT2_MAC = nn:nn:nn:nn:nn:nn
391 VANILLA_BRIDGE_IP = n.n.n.n
393 or use ``--test-params`` option
395 .. code-block:: console
397 $ ./vsperf --conf-file=<path_to_custom_conf>/10_custom.conf \
398 --test-params "VANILLA_TGEN_PORT1_IP=n.n.n.n;" \
399 "VANILLA_TGEN_PORT1_MAC=nn:nn:nn:nn:nn:nn;" \
400 "VANILLA_TGEN_PORT2_IP=n.n.n.n;" \
401 "VANILLA_TGEN_PORT2_MAC=nn:nn:nn:nn:nn:nn"
403 2. If needed, recompile src for all OVS variants
405 .. code-block:: console
413 .. code-block:: console
415 $ ./vsperf --conf-file<path_to_custom_conf>/10_custom.conf
422 Currently it is not possible to use standard scenario deployments for execution of
423 tests with VPP. It means, that deployments ``p2p``, ``pvp``, ``pvvp`` and in general any
424 :ref:`pxp-deployment` won't work with VPP. However it is possible to use VPP in
425 :ref:`step-driven-tests`. A basic set of VPP testcases covering ``phy2phy``, ``pvp``
426 and ``pvvp`` tests are already prepared.
428 List of performance tests with VPP support follows:
430 * phy2phy_tput_vpp: VPP: LTD.Throughput.RFC2544.PacketLossRatio
431 * phy2phy_cont_vpp: VPP: Phy2Phy Continuous Stream
432 * phy2phy_back2back_vpp: VPP: LTD.Throughput.RFC2544.BackToBackFrames
433 * pvp_tput_vpp: VPP: LTD.Throughput.RFC2544.PacketLossRatio
434 * pvp_cont_vpp: VPP: PVP Continuous Stream
435 * pvp_back2back_vpp: VPP: LTD.Throughput.RFC2544.BackToBackFrames
436 * pvvp_tput_vpp: VPP: LTD.Throughput.RFC2544.PacketLossRatio
437 * pvvp_cont_vpp: VPP: PVP Continuous Stream
438 * pvvp_back2back_vpp: VPP: LTD.Throughput.RFC2544.BackToBackFrames
440 In order to execute testcases with VPP it is required to:
442 * install VPP manually, see :ref:`vpp-installation`
443 * configure ``WHITELIST_NICS``, with two physical NICs connected to the traffic generator
444 * configure traffic generator, see :ref:`trafficgen-installation`
446 After that it is possible to execute VPP testcases listed above.
450 .. code-block:: console
452 $ ./vsperf --conf-file=<path_to_custom_conf> phy2phy_tput_vpp
456 Using vfio_pci with DPDK
457 ^^^^^^^^^^^^^^^^^^^^^^^^^
459 To use vfio with DPDK instead of igb_uio add into your custom configuration
460 file the following parameter:
462 .. code-block:: python
464 PATHS['dpdk']['src']['modules'] = ['uio', 'vfio-pci']
467 **NOTE:** In case, that DPDK is installed from binary package, then please
468 set ``PATHS['dpdk']['bin']['modules']`` instead.
470 **NOTE:** Please ensure that Intel VT-d is enabled in BIOS.
472 **NOTE:** Please ensure your boot/grub parameters include
475 **NOTE:** In case of VPP, it is required to explicitly define, that vfio-pci
476 DPDK driver should be used. It means to update dpdk part of VSWITCH_VPP_ARGS
477 dictionary with uio-driver section, e.g. VSWITCH_VPP_ARGS['dpdk'] = 'uio-driver vfio-pci'
479 .. code-block:: console
481 iommu=pt intel_iommu=on
483 To check that IOMMU is enabled on your platform:
485 .. code-block:: console
488 [ 0.000000] Intel-IOMMU: enabled
489 [ 0.139882] dmar: IOMMU 0: reg_base_addr fbffe000 ver 1:0 cap d2078c106f0466 ecap f020de
490 [ 0.139888] dmar: IOMMU 1: reg_base_addr ebffc000 ver 1:0 cap d2078c106f0466 ecap f020de
491 [ 0.139893] IOAPIC id 2 under DRHD base 0xfbffe000 IOMMU 0
492 [ 0.139894] IOAPIC id 0 under DRHD base 0xebffc000 IOMMU 1
493 [ 0.139895] IOAPIC id 1 under DRHD base 0xebffc000 IOMMU 1
494 [ 3.335744] IOMMU: dmar0 using Queued invalidation
495 [ 3.335746] IOMMU: dmar1 using Queued invalidation
503 To use virtual functions of NIC with SRIOV support, use extended form
504 of NIC PCI slot definition:
506 .. code-block:: python
508 WHITELIST_NICS = ['0000:05:00.0|vf0', '0000:05:00.1|vf3']
510 Where 'vf' is an indication of virtual function usage and following
511 number defines a VF to be used. In case that VF usage is detected,
512 then vswitchperf will enable SRIOV support for given card and it will
513 detect PCI slot numbers of selected VFs.
515 So in example above, one VF will be configured for NIC '0000:05:00.0'
516 and four VFs will be configured for NIC '0000:05:00.1'. Vswitchperf
517 will detect PCI addresses of selected VFs and it will use them during
520 At the end of vswitchperf execution, SRIOV support will be disabled.
522 SRIOV support is generic and it can be used in different testing scenarios.
525 * vSwitch tests with DPDK or without DPDK support to verify impact
526 of VF usage on vSwitch performance
527 * tests without vSwitch, where traffic is forwarded directly
528 between VF interfaces by packet forwarder (e.g. testpmd application)
529 * tests without vSwitch, where VM accesses VF interfaces directly
530 by PCI-passthrough_ to measure raw VM throughput performance.
534 Using QEMU with PCI passthrough support
535 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
537 Raw virtual machine throughput performance can be measured by execution of PVP
538 test with direct access to NICs by PCI pass-through. To execute VM with direct
539 access to PCI devices, enable vfio-pci_. In order to use virtual functions,
540 SRIOV-support_ must be enabled.
542 Execution of test with PCI pass-through with vswitch disabled:
544 .. code-block:: console
546 $ ./vsperf --conf-file=<path_to_custom_conf>/10_custom.conf \
547 --vswitch none --vnf QemuPciPassthrough pvp_tput
549 Any of supported guest-loopback-application_ can be used inside VM with
550 PCI pass-through support.
552 Note: Qemu with PCI pass-through support can be used only with PVP test
555 .. _guest-loopback-application:
557 Selection of loopback application for tests with VMs
558 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
560 To select the loopback applications which will forward packets inside VMs,
561 the following parameter should be configured:
563 .. code-block:: python
565 GUEST_LOOPBACK = ['testpmd']
567 or use ``--test-params`` CLI argument:
569 .. code-block:: console
571 $ ./vsperf --conf-file=<path_to_custom_conf>/10_custom.conf \
572 --test-params "GUEST_LOOPBACK=['testpmd']"
574 Supported loopback applications are:
576 .. code-block:: console
578 'testpmd' - testpmd from dpdk will be built and used
579 'l2fwd' - l2fwd module provided by Huawei will be built and used
580 'linux_bridge' - linux bridge will be configured
581 'buildin' - nothing will be configured by vsperf; VM image must
582 ensure traffic forwarding between its interfaces
584 Guest loopback application must be configured, otherwise traffic
585 will not be forwarded by VM and testcases with VM related deployments
586 will fail. Guest loopback application is set to 'testpmd' by default.
588 **NOTE:** In case that only 1 or more than 2 NICs are configured for VM,
589 then 'testpmd' should be used. As it is able to forward traffic between
590 multiple VM NIC pairs.
592 **NOTE:** In case of linux_bridge, all guest NICs are connected to the same
593 bridge inside the guest.
595 Mergable Buffers Options with QEMU
596 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
598 Mergable buffers can be disabled with VSPerf within QEMU. This option can
599 increase performance significantly when not using jumbo frame sized packets.
600 By default VSPerf disables mergable buffers. If you wish to enable it you
601 can modify the setting in the a custom conf file.
603 .. code-block:: python
605 GUEST_NIC_MERGE_BUFFERS_DISABLE = [False]
607 Then execute using the custom conf file.
609 .. code-block:: console
611 $ ./vsperf --conf-file=<path_to_custom_conf>/10_custom.conf
613 Alternatively you can just pass the param during execution.
615 .. code-block:: console
617 $ ./vsperf --test-params "GUEST_NIC_MERGE_BUFFERS_DISABLE=[False]"
620 Selection of dpdk binding driver for tests with VMs
621 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
623 To select dpdk binding driver, which will specify which driver the vm NICs will
624 use for dpdk bind, the following configuration parameter should be configured:
626 .. code-block:: console
628 GUEST_DPDK_BIND_DRIVER = ['igb_uio_from_src']
630 The supported dpdk guest bind drivers are:
632 .. code-block:: console
634 'uio_pci_generic' - Use uio_pci_generic driver
635 'igb_uio_from_src' - Build and use the igb_uio driver from the dpdk src
637 'vfio_no_iommu' - Use vfio with no iommu option. This requires custom
638 guest images that support this option. The default
639 vloop image does not support this driver.
641 Note: uio_pci_generic does not support sr-iov testcases with guests attached.
642 This is because uio_pci_generic only supports legacy interrupts. In case
643 uio_pci_generic is selected with the vnf as QemuPciPassthrough it will be
644 modified to use igb_uio_from_src instead.
646 Note: vfio_no_iommu requires kernels equal to or greater than 4.5 and dpdk
647 16.04 or greater. Using this option will also taint the kernel.
649 Please refer to the dpdk documents at http://dpdk.org/doc/guides for more
650 information on these drivers.
652 Guest Core and Thread Binding
653 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
655 VSPERF provides options to achieve better performance by guest core binding and
656 guest vCPU thread binding as well. Core binding is to bind all the qemu threads.
657 Thread binding is to bind the house keeping threads to some CPU and vCPU thread to
658 some other CPU, this helps to reduce the noise from qemu house keeping threads.
661 .. code-block:: python
663 GUEST_CORE_BINDING = [('#EVAL(6+2*#VMINDEX)', '#EVAL(7+2*#VMINDEX)')]
665 **NOTE** By default the GUEST_THREAD_BINDING will be none, which means same as
666 the GUEST_CORE_BINDING, i.e. the vcpu threads are sharing the physical CPUs with
667 the house keeping threads. Better performance using vCPU thread binding can be
668 achieved by enabling affinity in the custom configuration file.
670 For example, if an environment requires 32,33 to be core binded and 29,30&31 for
671 guest thread binding to achieve better performance.
673 .. code-block:: python
675 VNF_AFFINITIZATION_ON = True
676 GUEST_CORE_BINDING = [('32','33')]
677 GUEST_THREAD_BINDING = [('29', '30', '31')]
682 QEMU default to a compatible subset of performance enhancing cpu features.
683 To pass all available host processor features to the guest.
685 .. code-block:: python
687 GUEST_CPU_OPTIONS = ['host,migratable=off']
689 **NOTE** To enhance the performance, cpu features tsc deadline timer for guest,
690 the guest PMU, the invariant TSC can be provided in the custom configuration file.
692 Multi-Queue Configuration
693 ^^^^^^^^^^^^^^^^^^^^^^^^^
695 VSPerf currently supports multi-queue with the following limitations:
697 1. Requires QEMU 2.5 or greater and any OVS version higher than 2.5. The
698 default upstream package versions installed by VSPerf satisfies this
701 2. Guest image must have ethtool utility installed if using l2fwd or linux
702 bridge inside guest for loopback.
704 3. If using OVS versions 2.5.0 or less enable old style multi-queue as shown
705 in the ''02_vswitch.conf'' file.
707 .. code-block:: python
709 OVS_OLD_STYLE_MQ = True
711 To enable multi-queue for dpdk modify the ''02_vswitch.conf'' file.
713 .. code-block:: python
715 VSWITCH_DPDK_MULTI_QUEUES = 2
717 **NOTE:** you should consider using the switch affinity to set a pmd cpu mask
718 that can optimize your performance. Consider the numa of the NIC in use if this
719 applies by checking /sys/class/net/<eth_name>/device/numa_node and setting an
720 appropriate mask to create PMD threads on the same numa node.
722 When multi-queue is enabled, each dpdk or dpdkvhostuser port that is created
723 on the switch will set the option for multiple queues. If old style multi queue
724 has been enabled a global option for multi queue will be used instead of the
727 To enable multi-queue on the guest modify the ''04_vnf.conf'' file.
729 .. code-block:: python
731 GUEST_NIC_QUEUES = [2]
733 Enabling multi-queue at the guest will add multiple queues to each NIC port when
734 qemu launches the guest.
736 In case of Vanilla OVS, multi-queue is enabled on the tuntap ports and nic
737 queues will be enabled inside the guest with ethtool. Simply enabling the
738 multi-queue on the guest is sufficient for Vanilla OVS multi-queue.
740 Testpmd should be configured to take advantage of multi-queue on the guest if
741 using DPDKVhostUser. This can be done by modifying the ''04_vnf.conf'' file.
743 .. code-block:: python
745 GUEST_TESTPMD_PARAMS = ['-l 0,1,2,3,4 -n 4 --socket-mem 512 -- '
746 '--burst=64 -i --txqflags=0xf00 '
747 '--nb-cores=4 --rxq=2 --txq=2 '
750 **NOTE:** The guest SMP cores must be configured to allow for testpmd to use the
751 optimal number of cores to take advantage of the multiple guest queues.
753 In case of using Vanilla OVS and qemu virtio-net you can increase performance
754 by binding vhost-net threads to cpus. This can be done by enabling the affinity
755 in the ''04_vnf.conf'' file. This can be done to non multi-queue enabled
756 configurations as well as there will be 2 vhost-net threads.
758 .. code-block:: python
760 VSWITCH_VHOST_NET_AFFINITIZATION = True
762 VSWITCH_VHOST_CPU_MAP = [4,5,8,11]
764 **NOTE:** This method of binding would require a custom script in a real
767 **NOTE:** For optimal performance guest SMPs and/or vhost-net threads should be
768 on the same numa as the NIC in use if possible/applicable. Testpmd should be
769 assigned at least (nb_cores +1) total cores with the cpu mask.
774 VSPERF provides options to support jumbo frame testing with a jumbo frame supported
775 NIC and traffic generator for the following vswitches:
781 3. TestPMD loopback with or without a guest
783 **NOTE:** There is currently no support for SR-IOV or VPP at this time with jumbo
786 All packet forwarding applications for pxp testing is supported.
788 To enable jumbo frame testing simply enable the option in the conf files and set the
789 maximum size that will be used.
791 .. code-block:: python
793 VSWITCH_JUMBO_FRAMES_ENABLED = True
794 VSWITCH_JUMBO_FRAMES_SIZE = 9000
796 To enable jumbo frame testing with OVSVanilla the NIC in test on the host must have
797 its mtu size changed manually using ifconfig or applicable tools:
799 .. code-block:: console
801 ifconfig eth1 mtu 9000 up
803 **NOTE:** To make the setting consistent across reboots you should reference the OS
804 documents as it differs from distribution to distribution.
806 To start a test for jumbo frames modify the conf file packet sizes or pass the option
807 through the VSPERF command line.
809 .. code-block:: python
811 TEST_PARAMS = {'TRAFFICGEN_PKT_SIZES':(2000,9000)}
813 .. code-block:: python
815 ./vsperf --test-params "TRAFFICGEN_PKT_SIZES=2000,9000"
817 It is recommended to increase the memory size for OvsDpdkVhostUser testing from the default
818 1024. Your size required may vary depending on the number of guests in your testing. 4096
819 appears to work well for most typical testing scenarios.
821 .. code-block:: python
823 DPDK_SOCKET_MEM = ['4096', '0']
825 **NOTE:** For Jumbo frames to work with DpdkVhostUser, mergable buffers will be enabled by
826 default. If testing with mergable buffers in QEMU is desired, disable Jumbo Frames and only
827 test non jumbo frame sizes. Test Jumbo Frames sizes separately to avoid this collision.
830 Executing Packet Forwarding tests
831 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
833 To select the applications which will forward packets,
834 the following parameters should be configured:
836 .. code-block:: python
841 or use ``--vswitch`` and ``--fwdapp`` CLI arguments:
843 .. code-block:: console
845 $ ./vsperf phy2phy_cont --conf-file user_settings.py \
849 Supported Packet Forwarding applications are:
851 .. code-block:: console
853 'testpmd' - testpmd from dpdk
856 1. Update your ''10_custom.conf'' file to use the appropriate variables
857 for selected Packet Forwarder:
859 .. code-block:: python
861 # testpmd configuration
863 # packet forwarding mode supported by testpmd; Please see DPDK documentation
864 # for comprehensive list of modes supported by your version.
865 # e.g. io|mac|mac_retry|macswap|flowgen|rxonly|txonly|csum|icmpecho|...
866 # Note: Option "mac_retry" has been changed to "mac retry" since DPDK v16.07
867 TESTPMD_FWD_MODE = 'csum'
868 # checksum calculation layer: ip|udp|tcp|sctp|outer-ip
869 TESTPMD_CSUM_LAYER = 'ip'
870 # checksum calculation place: hw (hardware) | sw (software)
871 TESTPMD_CSUM_CALC = 'sw'
872 # recognize tunnel headers: on|off
873 TESTPMD_CSUM_PARSE_TUNNEL = 'off'
877 .. code-block:: console
879 $ ./vsperf phy2phy_tput --conf-file <path_to_settings_py>
881 Executing Packet Forwarding tests with one guest
882 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
884 TestPMD with DPDK 16.11 or greater can be used to forward packets as a switch to a single guest using TestPMD vdev
885 option. To set this configuration the following parameters should be used.
887 .. code-block:: python
892 or use ``--vswitch`` and ``--fwdapp`` CLI arguments:
894 .. code-block:: console
896 $ ./vsperf pvp_tput --conf-file user_settings.py \
900 Guest forwarding application only supports TestPMD in this configuration.
902 .. code-block:: python
904 GUEST_LOOPBACK = ['testpmd']
906 For optimal performance one cpu per port +1 should be used for TestPMD. Also set additional params for packet forwarding
907 application to use the correct number of nb-cores.
909 .. code-block:: python
911 DPDK_SOCKET_MEM = ['1024', '0']
912 VSWITCHD_DPDK_ARGS = ['-l', '46,44,42,40,38', '-n', '4']
913 TESTPMD_ARGS = ['--nb-cores=4', '--txq=1', '--rxq=1']
915 For guest TestPMD 3 VCpus should be assigned with the following TestPMD params.
917 .. code-block:: python
919 GUEST_TESTPMD_PARAMS = ['-l 0,1,2 -n 4 --socket-mem 1024 -- '
920 '--burst=64 -i --txqflags=0xf00 '
921 '--disable-hw-vlan --nb-cores=2 --txq=1 --rxq=1']
923 Execution of TestPMD can be run with the following command line
925 .. code-block:: console
927 ./vsperf pvp_tput --vswitch=none --fwdapp=TestPMD --conf-file <path_to_settings_py>
929 **NOTE:** To achieve the best 0% loss numbers with rfc2544 throughput testing, other tunings should be applied to host
930 and guest such as tuned profiles and CPU tunings to prevent possible interrupts to worker threads.
932 VSPERF modes of operation
933 ^^^^^^^^^^^^^^^^^^^^^^^^^
935 VSPERF can be run in different modes. By default it will configure vSwitch,
936 traffic generator and VNF. However it can be used just for configuration
937 and execution of traffic generator. Another option is execution of all
938 components except traffic generator itself.
940 Mode of operation is driven by configuration parameter -m or --mode
942 .. code-block:: console
944 -m MODE, --mode MODE vsperf mode of operation;
946 "normal" - execute vSwitch, VNF and traffic generator
947 "trafficgen" - execute only traffic generator
948 "trafficgen-off" - execute vSwitch and VNF
949 "trafficgen-pause" - execute vSwitch and VNF but wait before traffic transmission
951 In case, that VSPERF is executed in "trafficgen" mode, then configuration
952 of traffic generator can be modified through ``TRAFFIC`` dictionary passed to the
953 ``--test-params`` option. It is not needed to specify all values of ``TRAFFIC``
954 dictionary. It is sufficient to specify only values, which should be changed.
955 Detailed description of ``TRAFFIC`` dictionary can be found at
956 :ref:`configuration-of-traffic-dictionary`.
958 Example of execution of VSPERF in "trafficgen" mode:
960 .. code-block:: console
962 $ ./vsperf -m trafficgen --trafficgen IxNet --conf-file vsperf.conf \
963 --test-params "TRAFFIC={'traffic_type':'rfc2544_continuous','bidir':'False','framerate':60}"
968 The ``--matrix`` command line argument analyses and displays the performance of
969 all the tests run. Using the metric specified by ``MATRIX_METRIC`` in the conf-file,
970 the first test is set as the baseline and all the other tests are compared to it.
971 The ``MATRIX_METRIC`` must always refer to a numeric value to enable comparision.
972 A table, with the test ID, metric value, the change of the metric in %, testname
973 and the test parameters used for each test, is printed out as well as saved into the
976 Example of 2 tests being compared using Performance Matrix:
978 .. code-block:: console
980 $ ./vsperf --conf-file user_settings.py \
981 --test-params "['TRAFFICGEN_PKT_SIZES=(64,)',"\
982 "'TRAFFICGEN_PKT_SIZES=(128,)']" \
983 phy2phy_cont phy2phy_cont --matrix
987 .. code-block:: console
989 +------+--------------+---------------------+----------+---------------------------------------+
990 | ID | Name | throughput_rx_fps | Change | Parameters, CUMULATIVE_PARAMS = False |
991 +======+==============+=====================+==========+=======================================+
992 | 0 | phy2phy_cont | 23749000.000 | 0 | 'TRAFFICGEN_PKT_SIZES': [64] |
993 +------+--------------+---------------------+----------+---------------------------------------+
994 | 1 | phy2phy_cont | 16850500.000 | -29.048 | 'TRAFFICGEN_PKT_SIZES': [128] |
995 +------+--------------+---------------------+----------+---------------------------------------+
998 Code change verification by pylint
999 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1001 Every developer participating in VSPERF project should run
1002 pylint before his python code is submitted for review. Project
1003 specific configuration for pylint is available at 'pylint.rc'.
1005 Example of manual pylint invocation:
1007 .. code-block:: console
1009 $ pylint --rcfile ./pylintrc ./vsperf
1014 Custom image fails to boot
1015 ~~~~~~~~~~~~~~~~~~~~~~~~~~
1017 Using custom VM images may not boot within VSPerf pxp testing because of
1018 the drive boot and shared type which could be caused by a missing scsi
1019 driver inside the image. In case of issues you can try changing the drive
1022 .. code-block:: python
1024 GUEST_BOOT_DRIVE_TYPE = ['ide']
1025 GUEST_SHARED_DRIVE_TYPE = ['ide']
1027 OVS with DPDK and QEMU
1028 ~~~~~~~~~~~~~~~~~~~~~~~
1030 If you encounter the following error: "before (last 100 chars):
1031 '-path=/dev/hugepages,share=on: unable to map backing store for
1032 hugepages: Cannot allocate memory\r\n\r\n" during qemu initialization,
1033 check the amount of hugepages on your system:
1035 .. code-block:: console
1037 $ cat /proc/meminfo | grep HugePages
1040 By default the vswitchd is launched with 1Gb of memory, to change
1041 this, modify --socket-mem parameter in conf/02_vswitch.conf to allocate
1042 an appropriate amount of memory:
1044 .. code-block:: python
1046 DPDK_SOCKET_MEM = ['1024', '0']
1047 VSWITCHD_DPDK_ARGS = ['-c', '0x4', '-n', '4']
1048 VSWITCHD_DPDK_CONFIG = {
1049 'dpdk-init' : 'true',
1050 'dpdk-lcore-mask' : '0x4',
1051 'dpdk-socket-mem' : '1024,0',
1054 Note: Option ``VSWITCHD_DPDK_ARGS`` is used for vswitchd, which supports ``--dpdk``
1055 parameter. In recent vswitchd versions, option ``VSWITCHD_DPDK_CONFIG`` will be
1056 used to configure vswitchd via ``ovs-vsctl`` calls.
1062 For more information and details refer to the rest of vSwitchPerfuser documentation.