17. DPDK Test Suite

The DPDK Test Suite, abbreviated DTS, is a Python test framework with test suites implementing functional and performance tests used to test DPDK.

17.1. DTS Terminology

DTS node
A generic description of any host/server DTS connects to.
DTS runtime environment
An environment containing Python with packages needed to run DTS.
DTS runtime environment node
A node where at least one DTS runtime environment is present. This is the node where we run DTS and from which DTS connects to other nodes.
System under test
An SUT is the combination of DPDK and the hardware we’re testing in conjunction with DPDK (NICs, crypto and other devices).
System under test node
A node where at least one SUT is present.
Traffic generator
A TG is either software or hardware capable of sending packets.
Traffic generator node
A node where at least one TG is present. In case of hardware traffic generators, the TG and the node are literally the same.

In most cases, interchangeably referring to a runtime environment, SUT, TG or the node they’re running on (e.g. using SUT and SUT node interchangeably) doesn’t cause confusion. There could theoretically be more than of these running on the same node and in that case it’s useful to have stricter definitions. An example would be two different traffic generators (such as Trex and Scapy) running on the same node. A different example would be a node containing both a DTS runtime environment and a traffic generator, in which case it’s both a DTS runtime environment node and a TG node.

17.2. DTS Environment

DTS is written entirely in Python using a variety of dependencies. DTS uses Poetry as its Python dependency management. Python build/development and runtime environments are the same and DTS development environment, DTS runtime environment or just plain DTS environment are used interchangeably.

17.2.1. Setting up DTS environment

  1. Python Version

    The Python Version required by DTS is specified in dts/pyproject.toml in the [tool.poetry.dependencies] section:

    python = "^3.10"

    The Python dependency manager DTS uses, Poetry, doesn’t install Python, so you may need to satisfy this requirement by other means if your Python is not up-to-date. A tool such as Pyenv is a good way to get Python, though not the only one.

  2. Poetry

    The typical style of python dependency management, pip with requirements.txt, has a few issues. The advantages of Poetry include specifying what Python version is required and forcing you to specify versions, enforced by a lockfile, both of which help prevent broken dependencies. Another benefit is the usage of pyproject.toml, which has become the standard config file for python projects, improving project organization. To install Poetry, visit their doc pages. The recommended Poetry version is at least 1.5.1.

  3. Getting a Poetry shell

    Once you have Poetry along with the proper Python version all set up, it’s just a matter of installing dependencies via Poetry and using the virtual environment Poetry provides:

    poetry install --no-root
    poetry shell
  4. SSH Connection

    DTS uses the Fabric Python library for SSH connections between DTS environment and the other hosts. The authentication method used is pubkey authentication. Fabric tries to use a passed key/certificate, then any key it can with through an SSH agent, then any “id_rsa”, “id_dsa” or “id_ecdsa” key discoverable in ~/.ssh/ (with any matching OpenSSH-style certificates). DTS doesn’t pass any keys, so Fabric tries to use the other two methods.

17.3. Setting up System Under Test

There are two areas that need to be set up on a System Under Test:

  1. DPDK dependencies

    DPDK will be built and run on the SUT. Consult the Getting Started guides for the list of dependencies for each distribution.

  2. Hardware dependencies

    Any hardware DPDK uses needs a proper driver and most OS distributions provide those, but the version may not be satisfactory. It’s up to each user to install the driver they’re interested in testing. The hardware also may also need firmware upgrades, which is also left at user discretion.

  3. Hugepages

    There are two ways to configure hugepages:

    • DTS configuration

      You may specify the optional hugepage configuration in the DTS config file. If you do, DTS will take care of configuring hugepages, overwriting your current SUT hugepage configuration.

    • System under test configuration

      It’s possible to use the hugepage configuration already present on the SUT. If you wish to do so, don’t specify the hugepage configuration in the DTS config file.

  4. User with administrator privileges

DTS needs administrator privileges to run DPDK applications (such as testpmd) on the SUT. The SUT user must be able run commands in privileged mode without asking for password. On most Linux distributions, it’s a matter of setting up passwordless sudo:

  1. Run sudo visudo and check that it contains %sudo    ALL=(ALL:ALL) NOPASSWD:ALL.
  2. Add the SUT user to the sudo group with:
sudo usermod -aG sudo <sut_user>

17.4. Setting up Traffic Generator Node

These need to be set up on a Traffic Generator Node:

  1. Traffic generator dependencies

    The traffic generator running on the traffic generator node must be installed beforehand. For Scapy traffic generator, only a few Python libraries need to be installed:

    sudo apt install python3-pip
    sudo pip install --upgrade pip
    sudo pip install scapy==2.5.0
  2. Hardware dependencies

    The traffic generators, like DPDK, need a proper driver and firmware. The Scapy traffic generator doesn’t have strict requirements - the drivers that come with most OS distributions will be satisfactory.

  3. User with administrator privileges

    Similarly to the System Under Test, traffic generators need administrator privileges to be able to use the devices. Refer to the System Under Test section <sut_admin_user> for details.

17.5. Running DTS

DTS needs to know which nodes to connect to and what hardware to use on those nodes. Once that’s configured, either a DPDK source code tarball or a Git revision ID of choice needs to be supplied. DTS will use this to compile DPDK on the SUT node and then run the tests with the newly built binaries.

17.5.1. Configuring DTS

DTS configuration is split into nodes and executions and build targets within executions, and follows a defined schema as described in Configuration Schema. By default, DTS will try to use the dts/conf.yaml config file, which is a template that illustrates what can be configured in DTS.

The user must have administrator privileges which don’t require password authentication.

17.5.2. DTS Execution

DTS is run with main.py located in the dts directory after entering Poetry shell:

(dts-py3.10) $ ./main.py --help
usage: main.py [-h] [--config-file CONFIG_FILE] [--output-dir OUTPUT_DIR] [-t TIMEOUT] [-v] [-s] [--tarball TARBALL] [--compile-timeout COMPILE_TIMEOUT] [--test-suite TEST_SUITE [TEST_CASES ...]] [--re-run RE_RUN]

Run DPDK test suites. All options may be specified with the environment variables provided in brackets. Command line arguments have higher priority.

-h, --help            show this help message and exit
--config-file CONFIG_FILE
                      [DTS_CFG_FILE] The configuration file that describes the test cases, SUTs and targets. (default: ./conf.yaml)
--output-dir OUTPUT_DIR, --output OUTPUT_DIR
                      [DTS_OUTPUT_DIR] Output directory where DTS logs and results are saved. (default: output)
-t TIMEOUT, --timeout TIMEOUT
                      [DTS_TIMEOUT] The default timeout for all DTS operations except for compiling DPDK. (default: 15)
-v, --verbose         [DTS_VERBOSE] Specify to enable verbose output, logging all messages to the console. (default: False)
-s, --skip-setup      [DTS_SKIP_SETUP] Specify to skip all setup steps on SUT and TG nodes. (default: False)
--tarball TARBALL, --snapshot TARBALL, --git-ref TARBALL
                      [DTS_DPDK_TARBALL] Path to DPDK source code tarball or a git commit ID, tag ID or tree ID to test. To test local changes, first commit them, then use the commit ID with this option. (default: dpdk.tar.xz)
--compile-timeout COMPILE_TIMEOUT
                      [DTS_COMPILE_TIMEOUT] The timeout for compiling DPDK. (default: 1200)
--test-suite TEST_SUITE [TEST_CASES ...]
                      [DTS_TEST_SUITES] A list containing a test suite with test cases. The first parameter is the test suite name, and the rest are test case names, which are optional. May be specified multiple times. To specify multiple test suites in the environment
                      variable, join the lists with a comma. Examples: --test-suite suite case case --test-suite suite case ... | DTS_TEST_SUITES='suite case case, suite case, ...' | --test-suite suite --test-suite suite case ... | DTS_TEST_SUITES='suite, suite case, ...'
                      (default: [])
--re-run RE_RUN, --re_run RE_RUN
                      [DTS_RERUN] Re-run each test case the specified number of times if a test failure occurs. (default: 0)

The brackets contain the names of environment variables that set the same thing. The minimum DTS needs is a config file and a DPDK tarball or git ref ID. You may pass those to DTS using the command line arguments or use the default paths.

Example command for running DTS with the template configuration and DPDK tag v23.11:

(dts-py3.10) $ ./main.py --git-ref v23.11

17.5.3. DTS Results

Results are stored in the output dir by default which be changed with the --output-dir command line argument. The results contain basic statistics of passed/failed test cases and DPDK version.

17.6. Contributing to DTS

There are two areas of contribution: The DTS framework and DTS test suites.

The framework contains the logic needed to run test cases, such as connecting to nodes, running DPDK applications and collecting results.

The test cases call APIs from the framework to test their scenarios. Adding test cases may require adding code to the framework as well.

17.6.1. Framework Coding Guidelines

When adding code to the DTS framework, pay attention to the rest of the code and try not to divert much from it. The DTS developer tools will issue warnings when some of the basics are not met.

The code must be properly documented with docstrings. The style must conform to the Google style. See an example of the style here. For cases which are not covered by the Google style, refer to PEP 257. There are some cases which are not covered by the two style guides, where we deviate or where some additional clarification is helpful:

  • The __init__() methods of classes are documented separately from the docstring of the class itself.

  • The docstrings of implemented abstract methods should refer to the superclass’s definition if there’s no deviation.

  • Instance variables/attributes should be documented in the docstring of the class in the Attributes: section.

  • The dataclass.dataclass decorator changes how the attributes are processed. The dataclass attributes which result in instance variables/attributes should also be recorded in the Attributes: section.

  • Class variables/attributes, on the other hand, should be documented with #: above the type annotated line. The description may be omitted if the meaning is obvious.

  • The Enum and TypedDict also process the attributes in particular ways and should be documented with #: as well. This is mainly so that the autogenerated documentation contains the assigned value.

  • When referencing a parameter of a function or a method in their docstring, don’t use any articles and put the parameter into single backticks. This mimics the style of Python’s documentation.

  • When specifying a value, use double backticks:

    def foo(greet: bool) -> None:
        """Demonstration of single and double backticks.
        `greet` controls whether ``Hello World`` is printed.
           greet: Whether to print the ``Hello World`` message.
        if greet:
           print(f"Hello World")
  • The docstring maximum line length is the same as the code maximum line length.

17.7. How To Write a Test Suite

All test suites inherit from TestSuite defined in dts/framework/test_suite.py. There are four types of methods that comprise a test suite:

  1. Test cases

    Test cases are methods that start with a particular prefix.
    Functional test cases start with test_, e.g. test_hello_world_single_core.
    Performance test cases start with test_perf_, e.g. test_perf_nic_single_core.
    A test suite may have any number of functional and/or performance test cases. However, these test cases must test the same feature, following the rule of one feature = one test suite. Test cases for one feature don’t need to be grouped in just one test suite, though. If the feature requires many testing scenarios to cover, the test cases would be better off spread over multiple test suites so that each test suite doesn’t take too long to execute.
  2. Setup and Teardown methods

    There are setup and teardown methods for the whole test suite and each individual test case.
    Methods set_up_suite and tear_down_suite will be executed before any and after all test cases have been executed, respectively.
    Methods set_up_test_case and tear_down_test_case will be executed before and after each test case, respectively.
    These methods don’t need to be implemented if there’s no need for them in a test suite. In that case, nothing will happen when they are executed.
  3. Configuration, traffic and other logic

    The TestSuite class contains a variety of methods for anything that a test suite setup, a teardown, or a test case may need to do.

    The test suites also frequently use a DPDK app, such as testpmd, in interactive mode and use the interactive shell instances directly.

    These are the two main ways to call the framework logic in test suites. If there’s any functionality or logic missing from the framework, it should be implemented so that the test suites can use one of these two ways.

  4. Test case verification

    Test case verification should be done with the verify method, which records the result. The method should be called at the end of each test case.

  5. Other methods

    Of course, all test suite code should adhere to coding standards. Only the above methods will be treated specially and any other methods may be defined (which should be mostly private methods needed by each particular test suite). Any specific features (such as NIC configuration) required by a test suite should be implemented in the SutNode class (and the underlying classes that SutNode uses) and used by the test suite via the sut_node field.

17.8. DTS Developer Tools

There are three tools used in DTS to help with code checking, style and formatting:

  • isort

    Alphabetically sorts python imports within blocks.

  • black

    Does most of the actual formatting (whitespaces, comments, line length etc.) and works similarly to clang-format.

  • pylama

    Runs a collection of python linters and aggregates output. It will run these tools over the repository:

    linters = "mccabe,pycodestyle,pydocstyle,pyflakes"
  • mypy

    Enables static typing for Python, exploiting the type hints in the source code.

These three tools are all used in devtools/dts-check-format.sh, the DTS code check and format script. Refer to the script for usage: devtools/dts-check-format.sh -h.

17.9. Configuration Schema

17.9.1. Definitions

Node name
string – A unique identifier for a node. Examples: SUT1, TG1.
string – The CPU architecture. Supported values: x86_64, arm64, ppc64le.
string – The CPU microarchitecture. Use native for x86. Supported values: native, armv8a, dpaa2, thunderx, xgene1.
string – The operating system. Supported values: linux.
string – The compiler used for building DPDK. Supported values: gcc, clang, icc, mscv.
Build target

mapping – Build targets supported by DTS for building DPDK, described as:

arch See ARCH
os See OS
cpu See CPU
compiler See Compiler

string – Value prepended to the CC variable for the DPDK build.

Example: ccache


mapping – hugepages described as:


integer – The amount of hugepages to configure.

Hugepage size will be the system default.


(optional, defaults to false) – If true, it forces the

configuration of hugepages on the first NUMA node.

Network port

mapping – the NIC port described as:

pci string – the local PCI address of the port. Example: 0000:00:08.0
string – this port’s device driver when using with DPDK
When setting up the SUT, DTS will bind the network device to this driver
for compatibility with DPDK.

Examples: vfio-pci, mlx5_core

string – this port’s device driver when not using with DPDK
When tearing down the tests on the SUT, DTS will bind the network device
back to this driver. This driver is meant to be the one that the SUT would
normally use for this device, or whichever driver it is preferred to leave the
device bound to after testing.
This also represents the driver that is used in conjunction with the traffic
generator software.

Examples: i40e, mlx5_core

peer_node string – the name of the peer node connected to this port.
peer_pci string – the PCI address of the peer node port. Example: 000a:01:00.1
Test suite
string – name of the test suite to run. Examples: hello_world, os_udp
Test target

mapping – selects specific test cases to run from a test suite. Mapping is described as follows:

suite See Test suite

(optional) sequence of string – list of the selected test cases in the test suite to run.

Unknown test cases will be silently ignored.

17.9.2. Properties

The configuration requires listing all the execution environments and nodes involved in the testing. These can be defined with the following mappings:


sequence listing the execution environments. Each entry is described as per the following mapping:

build_targets sequence of Build target
perf boolean – Enable performance testing.
func boolean – Enable functional testing.
test_suites sequence of one of Test suite or Test target
skip_smoke_tests (optional) boolean – Allows you to skip smoke testing if true.
system_under_test_node System under test node specified with:
node_name See Node name

(optional) sequence of string

List of virtual devices passed with the --vdev argument to DPDK. Example: crypto_openssl

traffic_generator_node Node name for the traffic generator node.

sequence listing the nodes. Each entry is described as per the following mapping:

name See Node name
hostname string – The network hostname or IP address of this node.
user string – The SSH user credential to use to login to this node.

(optional) string – The SSH password credential for this node.

NB: Use only as last resort. SSH keys are strongly preferred.

arch The architecture of this node. See ARCH for supported values.
os The operating system of this node. See OS for supported values.
(optional, defaults to 1) string – Comma-separated list of logical
cores to use. An empty string means use all lcores.

Example: 1,2,3,4,5,18-22


(optional, defaults to false) boolean

Indicates whether DPDK should use only the first physical core or not.


(optional, defaults to 1) integer

The number of the memory channels to use.


(optional) See hugepages. If unset, hugepages won’t be configured

in favour of the system configuration.

sequence of Network port – Describe ports that are directly paired with
other nodes used in conjunction with this one. Both ends of the links must be
described. If there any inconsistencies DTS won’t run.

Example: port 1 of node SUT1 is connected to port 1 of node TG1 etc.

traffic_generator (optional) Traffic generator, if any, setup on this node described as:
type stringSupported values: SCAPY

17.9.3. Example

The following example (which can be found in dts/conf.yaml) sets up two nodes:

  • SUT1 which is already setup with the DPDK build requirements and any other required for execution;
  • TG1 which already has Scapy installed in the system.

And they both have two network ports which are physically connected to each other.


This example assumes that you have setup SSH keys in both the system under test and traffic generator nodes.

  # define one execution environment
  - build_targets:
      - arch: x86_64
        os: linux
        cpu: native
        # the combination of the following two makes CC="ccache gcc"
        compiler: gcc
        compiler_wrapper: ccache
    perf: false # disable performance testing
    func: true # enable functional testing
    skip_smoke_tests: false # optional
    test_suites: # the following test suites will be run in their entirety
      - hello_world
      - os_udp
    # The machine running the DPDK test executable
      node_name: "SUT 1"
      vdevs: # optional; if removed, vdevs won't be used in the execution
        - "crypto_openssl"
    # Traffic generator node to use for this execution environment
    traffic_generator_node: "TG 1"
  # Define a system under test node, having two network ports physically
  # connected to the corresponding ports in TG 1 (the peer node)
  - name: "SUT 1"
    hostname: sut1.change.me.localhost
    user: dtsuser
    arch: x86_64
    os: linux
    lcores: "" # use all the available logical cores
    use_first_core: false # tells DPDK to use any physical core
    memory_channels: 4 # tells DPDK to use 4 memory channels
    hugepages:  # optional; if removed, will use system hugepage configuration
        amount: 256
        force_first_numa: false
      # sets up the physical link between "SUT 1"@000:00:08.0 and "TG 1"@0000:00:08.0
      - pci: "0000:00:08.0"
        os_driver_for_dpdk: vfio-pci # OS driver that DPDK will use
        os_driver: i40e              # OS driver to bind when the tests are not running
        peer_node: "TG 1"
        peer_pci: "0000:00:08.0"
      # sets up the physical link between "SUT 1"@000:00:08.1 and "TG 1"@0000:00:08.1
      - pci: "0000:00:08.1"
        os_driver_for_dpdk: vfio-pci
        os_driver: i40e
        peer_node: "TG 1"
        peer_pci: "0000:00:08.1"
  # Define a Scapy traffic generator node, having two network ports
  # physically connected to the corresponding ports in SUT 1 (the peer node).
  - name: "TG 1"
    hostname: tg1.change.me.localhost
    user: dtsuser
    arch: x86_64
    os: linux
      # sets up the physical link between "TG 1"@000:00:08.0 and "SUT 1"@0000:00:08.0
      - pci: "0000:00:08.0"
        os_driver_for_dpdk: rdma
        os_driver: rdma
        peer_node: "SUT 1"
        peer_pci: "0000:00:08.0"
      # sets up the physical link between "SUT 1"@000:00:08.0 and "TG 1"@0000:00:08.0
      - pci: "0000:00:08.1"
        os_driver_for_dpdk: rdma
        os_driver: rdma
        peer_node: "SUT 1"
        peer_pci: "0000:00:08.1"
    hugepages:  # optional; if removed, will use system hugepage configuration
        amount: 256
        force_first_numa: false
        type: SCAPY