ceph/src/spdk/dpdk/doc/guides/linux_gsg/nic_perf_intel_platform.rst

   1 How to get best performance with NICs on Intel platforms
   2 ========================================================
   3
   4 This document is a step-by-step guide for getting high performance from DPDK applications on Intel platforms.
   5
   6
   7 Hardware and Memory Requirements
   8 --------------------------------
   9
  10 For best performance use an Intel Xeon class server system such as Ivy Bridge, Haswell or newer.
  11
  12 Ensure that each memory channel has at least one memory DIMM inserted, and that the memory size for each is at least 4GB.
  13 **Note**: this has one of the most direct effects on performance.
  14
  15 You can check the memory configuration using ``dmidecode`` as follows::
  16
  17       dmidecode -t memory | grep Locator
  18
  19       Locator: DIMM_A1
  20       Bank Locator: NODE 1
  21       Locator: DIMM_A2
  22       Bank Locator: NODE 1
  23       Locator: DIMM_B1
  24       Bank Locator: NODE 1
  25       Locator: DIMM_B2
  26       Bank Locator: NODE 1
  27       ...
  28       Locator: DIMM_G1
  29       Bank Locator: NODE 2
  30       Locator: DIMM_G2
  31       Bank Locator: NODE 2
  32       Locator: DIMM_H1
  33       Bank Locator: NODE 2
  34       Locator: DIMM_H2
  35       Bank Locator: NODE 2
  36
  37 The sample output above shows a total of 8 channels, from ``A`` to ``H``, where each channel has 2 DIMMs.
  38
  39 You can also use ``dmidecode`` to determine the memory frequency::
  40
  41       dmidecode -t memory | grep Speed
  42
  43       Speed: 2133 MHz
  44       Configured Clock Speed: 2134 MHz
  45       Speed: Unknown
  46       Configured Clock Speed: Unknown
  47       Speed: 2133 MHz
  48       Configured Clock Speed: 2134 MHz
  49       Speed: Unknown
  50       ...
  51       Speed: 2133 MHz
  52       Configured Clock Speed: 2134 MHz
  53       Speed: Unknown
  54       Configured Clock Speed: Unknown
  55       Speed: 2133 MHz
  56       Configured Clock Speed: 2134 MHz
  57       Speed: Unknown
  58       Configured Clock Speed: Unknown
  59
  60 The output shows a speed of 2133 MHz (DDR4) and Unknown (not existing).
  61 This aligns with the previous output which showed that each channel has one memory bar.
  62
  63
  64 Network Interface Card Requirements
  65 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  66
  67 Use a `DPDK supported <http://core.dpdk.org/supported/>`_ high end NIC such as the Intel XL710 40GbE.
  68
  69 Make sure each NIC has been flashed the latest version of NVM/firmware.
  70
  71 Use PCIe Gen3 slots, such as Gen3 ``x8`` or Gen3 ``x16`` because PCIe Gen2 slots don't provide enough bandwidth
  72 for 2 x 10GbE and above.
  73 You can use ``lspci`` to check the speed of a PCI slot using something like the following::
  74
  75       lspci -s 03:00.1 -vv | grep LnkSta
  76
  77       LnkSta: Speed 8GT/s, Width x8, TrErr- Train- SlotClk+ DLActive- ...
  78       LnkSta2: Current De-emphasis Level: -6dB, EqualizationComplete+ ...
  79
  80 When inserting NICs into PCI slots always check the caption, such as CPU0 or CPU1 to indicate which socket it is connected to.
  81
  82 Care should be take with NUMA.
  83 If you are using 2 or more ports from different NICs, it is best to ensure that these NICs are on the same CPU socket.
  84 An example of how to determine this is shown further below.
  85
  86
  87 BIOS Settings
  88 ~~~~~~~~~~~~~
  89
  90 The following are some recommendations on BIOS settings. Different platforms will have different BIOS naming
  91 so the following is mainly for reference:
  92
  93 #. Establish the steady state for the system, consider reviewing BIOS settings desired for best performance characteristic e.g. optimize for performance or energy efficiency.
  94
  95 #. Match the BIOS settings to the needs of the application you are testing.
  96
  97 #. Typically, **Performance** as the CPU Power and Performance policy is a reasonable starting point.
  98
  99 #. Consider using Turbo Boost to increase the frequency on cores.
 100
 101 #. Disable all virtualization options when you test the physical function of the NIC, and turn on VT-d if you wants to use VFIO.
 102
 103
 104 Linux boot command line
 105 ~~~~~~~~~~~~~~~~~~~~~~~
 106
 107 The following are some recommendations on GRUB boot settings:
 108
 109 #. Use the default grub file as a starting point.
 110
 111 #. Reserve 1G huge pages via grub configurations. For example to reserve 8 huge pages of 1G size::
 112
 113       default_hugepagesz=1G hugepagesz=1G hugepages=8
 114
 115 #. Isolate CPU cores which will be used for DPDK. For example::
 116
 117       isolcpus=2,3,4,5,6,7,8
 118
 119 #. If it wants to use VFIO, use the following additional grub parameters::
 120
 121       iommu=pt intel_iommu=on
 122
 123
 124 Configurations before running DPDK
 125 ----------------------------------
 126
 127 1. Build the DPDK target and reserve huge pages.
 128    See the earlier section on :ref:`linux_gsg_hugepages` for more details.
 129
 130    The following shell commands may help with building and configuration:
 131
 132    .. code-block:: console
 133
 134       # Build DPDK target.
 135       cd dpdk_folder
 136       make install T=x86_64-native-linux-gcc -j
 137
 138       # Get the hugepage size.
 139       awk '/Hugepagesize/ {print $2}' /proc/meminfo
 140
 141       # Get the total huge page numbers.
 142       awk '/HugePages_Total/ {print $2} ' /proc/meminfo
 143
 144       # Unmount the hugepages.
 145       umount `awk '/hugetlbfs/ {print $2}' /proc/mounts`
 146
 147       # Create the hugepage mount folder.
 148       mkdir -p /mnt/huge
 149
 150       # Mount to the specific folder.
 151       mount -t hugetlbfs nodev /mnt/huge
 152
 153 2. Check the CPU layout using using the DPDK ``cpu_layout`` utility:
 154
 155    .. code-block:: console
 156
 157       cd dpdk_folder
 158
 159       usertools/cpu_layout.py
 160
 161    Or run ``lscpu`` to check the cores on each socket.
 162
 163 3. Check your NIC id and related socket id:
 164
 165    .. code-block:: console
 166
 167       # List all the NICs with PCI address and device IDs.
 168       lspci -nn | grep Eth
 169
 170    For example suppose your output was as follows::
 171
 172       82:00.0 Ethernet [0200]: Intel XL710 for 40GbE QSFP+ [8086:1583]
 173       82:00.1 Ethernet [0200]: Intel XL710 for 40GbE QSFP+ [8086:1583]
 174       85:00.0 Ethernet [0200]: Intel XL710 for 40GbE QSFP+ [8086:1583]
 175       85:00.1 Ethernet [0200]: Intel XL710 for 40GbE QSFP+ [8086:1583]
 176
 177    Check the PCI device related numa node id:
 178
 179    .. code-block:: console
 180
 181       cat /sys/bus/pci/devices/0000\:xx\:00.x/numa_node
 182
 183    Usually ``0x:00.x`` is on socket 0 and ``8x:00.x`` is on socket 1.
 184    **Note**: To get the best performance, ensure that the core and NICs are in the same socket.
 185    In the example above ``85:00.0`` is on socket 1 and should be used by cores on socket 1 for the best performance.
 186
 187 4. Check which kernel drivers needs to be loaded and whether there is a need to unbind the network ports from their kernel drivers.
 188 More details about DPDK setup and Linux kernel requirements see :ref:`linux_gsg_compiling_dpdk` and :ref:`linux_gsg_linux_drivers`.