Intel XL710-Q2 User manual

Intel®Ethernet Adapters and Devices
User Guide

Overview
Welcome to the User's Guide for Intel® Ethernet Adapters and devices. This guide covers hardware and
software installation, setup procedures, and troubleshooting tips for Intel network adapters, connections, and
other devices.
Installing the Network Adapter
If you are installing a network adapter, follow this procedure from step 1.
If you are upgrading the driver software, start with step 5 .
1. Make sure that you are installing the latest driver software for your adapter. Visit Intel's support web-
site to download the latest drivers.
2. Review system requirements.
3. Insert the adapter(s) in the computer.
4. Attach the copper or fiber network cable(s).
5. Install the driver.
6. For Windows systems, install the Intel® PROSet software.
If you have any problems with basic installation, see Troubleshooting.
You can now set up advanced features, if necessary. The available features and the configuration process
varies with the adapter and your operating system.
Before You Begin
Supported Devices
For help identifying your network device and finding supported devices, click the link below:
http://www.intel.com/support
Compatibility Notes
In order for an adapter based on the XL710 controller to reach its full potential, you must install it in a PCIe
Gen3 x8 slot. Installing it in a shorter slot, or a Gen2 or Gen1 slot, will limit the throughput of the adapter.
Some older Intel® Ethernet Adapters do not have full software support for the most recent versions of
Microsoft Windows*. Many older Intel Ethernet Adapters have base drivers supplied by Microsoft Windows.
Lists of supported devices per OS are available at
http://www.intel.com/support/go/network/adapter/nicoscomp.htm
Supported Operating Systems
Supported 32-bit Operating Systems
NOTE: Microsoft* Windows* 32-bit operating systems are only supported on Intel 1GbE Ethernet
Adapters and slower devices. All adapters support 32-bit versions of Linux* and FreeBSD*.
Basic software and drivers are supported on the following operating systems:
lDOS
lSunSoft* Solaris* (drivers and support are provided by the operating system vendor)

Advanced software and drivers are supported on the following operating systems:
lMicrosoft Windows 7
lMicrosoft Windows 8
lMicrosoft Windows 8.1
lMicrosoft Windows 10
lLinux*, v2.4 kernel or higher
lFreeBSD*
Supported Intel® 64 Architecture Operating Systems
lMicrosoft* Windows* 7
lMicrosoft Windows 8
lMicrosoft Windows 8.1
lMicrosoft Windows 10
lMicrosoft* Windows Server* 2008 R2
lMicrosoft Windows Server 2012
lMicrosoft Windows Server 2012 R2
lMicrosoft Windows Server 2016
lMicrosoft Windows Server 2016 Nano Server
lVMWare ESXi 5.5
lVMWare* ESXi* 6.0
lVMWare ESXi 6.5
lRed Hat* Linux*
lNovell* SUSE* Linux
lFreeBSD*
Supported Operating Systems for Itanium-based Systems
lLinux, v2.x kernel and higher, except v2.6
Some older Intel® Ethernet Adapters do not have full software support for the most recent versions of
Microsoft Windows*. Many older Intel Ethernet Adapters have base drivers supplied by Microsoft Windows.
Lists of supported devices per OS are available at
http://www.intel.com/support/go/network/adapter/nicoscomp.htm
Hardware Compatibility
Before installing the adapter, check your system for the following:
lThe latest BIOS for your system
lOne open PCI Express slot
NOTE: The Intel® 10 Gigabit AT Server Adapter will only fit into x8 or larger PCI Express slots.
Some systems have physical x8 PCI Express slots that actually support lower speeds. Please
check your system manual to identify the slot.

Cabling Requirements
Intel Gigabit Adapters
Fiber Optic Cables
lLaser wavelength: 850 nanometer (not visible).
lSC Cable type:
lMulti-mode fiber with 50 micron core diameter; maximum length is 550 meters.
lMulti-mode fiber with 62.5 micron core diameter; maximum length is 275 meters.
lConnector type: SC.
lLC Cable type:
lMulti-mode fiber with 50 micron core diameter; maximum length is 550 meters.
lMulti-mode fiber with 62.5 micron core diameter; maximum length is 275 meters.
lConnector type: LC.
Copper Cables
l1000BASE-T or 100BASE-TX on Category 5 or Category 5e wiring, twisted 4-pair copper:
lMake sure you use Category 5 cabling that complies with the TIA-568 wiring specification. For
more information on this specification, see the Telecommunications Industry Association's web
site: www.tiaonline.org.
lMaximum Length is 100 meters.
lCategory 3 wiring supports only 10 Mbps.
NOTE: To insure compliance with CISPR 24 and the EU’s EN55024, devices based on the 82576
controller should be used only with CAT 5E shielded cables that are properly terminated according
to the recommendations in EN50174-2.
Intel 10 Gigabit Adapters
Fiber Optic Cables
lLaser wavelength: 850 nanometer (not visible).
lSC Cable type:
lMulti-mode fiber with 50 micron core diameter; maximum length is 550 meters.
lMulti-mode fiber with 62.5 micron core diameter; maximum length is 275 meters.
lConnector type: SC.
lLC Cable type:
lMulti-mode fiber with 50 micron core diameter; maximum length is 550 meters.
lMulti-mode fiber with 62.5 micron core diameter; maximum length is 275 meters.
lConnector type: LC.
Copper Cables
lMaximum lengths for Intel® 10 Gigabit Server Adapters and Connections that use 10GBASE-T on Cat-
egory 6, Category 6a, or Category 7 wiring, twisted 4-pair copper:
lMaximum length for Category 6 is 55 meters.
lMaximum length for Category 6a is 100 meters.
lMaximum length for Category 7 is 100 meters.

lTo ensure compliance with CISPR 24 and the EU's EN55024, Intel® 10 Gigabit Server
Adapters and Connections should be used only with CAT 6a shielded cables that are properly
terminated according to the recommendations in EN50174-2.
l10 Gigabit Ethernet over SFP+ Direct Attached Cable (Twinaxial)
lLength is 10 meters max.
Intel 40 Gigabit Adapters
Fiber Optic Cables
lLaser wavelength: 850 nanometer (not visible).
lSC Cable type:
lMulti-mode fiber with 50 micron core diameter; maximum length is 550 meters.
lMulti-mode fiber with 62.5 micron core diameter; maximum length is 275 meters.
lConnector type: SC.
lLC Cable type:
lMulti-mode fiber with 50 micron core diameter; maximum length is 550 meters.
lMulti-mode fiber with 62.5 micron core diameter; maximum length is 275 meters.
lConnector type: LC.
Copper Cables
l40 Gigabit Ethernet over SFP+ Direct Attached Cable (Twinaxial)
lLength is 7 meters max
Installation Overview
Installing the Adapter
1. Turn off the computer and unplug the power cord.
2. Remove the computer cover and the adapter slot cover from the slot that matches your adapter.
3. Insert the adapter edge connector into the slot and secure the bracket to the chassis.
4. Replace the computer cover, then plug in the power cord.
Install Drivers and Software
Windows* Operating Systems
You must have administrative rights to the operating system to install the drivers.
1. Download the latest drivers from the support website and transfer them to the system.
2. If the Found New Hardware Wizard screen is displayed, click Cancel.
3. Start the autorun located in the downloaded the software package. The autorun may automatically start
after you have extracted the files.
4. Click Install Drivers and Software.
5. Follow the instructions in the install wizard.

Installing Linux* Drivers from Source Code
1. Download and expand the base driver tar file.
2. Compile the driver module.
3. Install the module using the modprobe command.
4. Assign an IP address using the ifconfig command.
Optimizing Performance
You can configure Intel network adapter advanced settings to help optimize server performance.
The examples below provide guidance for three server usage models:
lOptimized for quick response and low latency – useful for video, audio, and High Performance Com-
puting Cluster (HPCC) servers
lOptimized for throughput – useful for data backup/retrieval and file servers
lOptimized for CPU utilization – useful for application, web, mail, and database servers
NOTES:
lThe recommendations below are guidelines and should be treated as such. Additional factors
such as installed applications, bus type, network topology, and operating system also affect
system performance.
lThese adjustments should be performed by a highly skilled network administrator. They are
not guaranteed to improve performance. Not all settings shown here may be available
through network driver configuration, operating system or system BIOS. Linux users, see the
README file in the Linux driver package for Linux-specific performance enhancement
details.
lWhen using performance test software, refer to the documentation of the application for
optimal results.
General Optimization
lInstall the adapter in an appropriate slot.
NOTE: Some PCIe x8 slots are actually configured as x4 slots. These slots have insufficient
bandwidth for full line rate with some dual port devices. The driver can detect this situation
and will write the following message in the system log: “PCI-Express bandwidth available for
this card is not sufficient for optimal performance. For optimal performance a x8 PCI-
Express slot is required.”If this error occurs, moving your adapter to a true x8 slot will resolve
the issue.
lIn order for an Intel® X710/XL710 based Network Adapter to reach its full potential, you must install it
in a PCIe Gen3 x8 slot. Installing it in a shorter slot, or a Gen2 or Gen1 slot, will impact the throughput
the adapter can attain.
lUse the proper cabling for your device.
lEnable Jumbo Packets, if your other network components can also be configured for it.
lIncrease the number of TCP and Socket resources from the default value. For Windows based sys-
tems, we have not identified system parameters other than the TCP Window Size which significantly
impact performance.

lIncrease the allocation size of Driver Resources (transmit/receive buffers). However, most TCP traffic
patterns work best with the transmit buffer set to its default value, and the receive buffer set to its min-
imum value.
lWhen passing traffic on multiple network ports using an I/O application that runs on most or all of the
cores in your system, consider setting the CPU Affinity for that application to fewer cores. This should
reduce CPU utilization and in some cases may increase throughput for the device. The cores selected
for CPU Affinity must be local to the affected network device's Processor Node/Group. You can use
the PowerShell command Get-NetAdapterRSS to list the cores that are local to a device. You may
need to increase the number of cores assigned to the application to maximize throughput. Refer to your
operating system documentation for more details on setting the CPU Affinity.
lIf you have multiple 10 Gpbs (or faster) ports installed in a system, the RSS queues of each adapter
port can be adjusted to use non-overlapping sets of processors within the adapter's local NUMA
Node/Socket. Change the RSS Base Processor Number for each adapter port so that the combination
of the base processor and the max number of RSS processors settings ensure non-overlapping cores.
1. Identify the adapter ports to be adjusted and inspect at their RssProcessorArray using the
Get-NetAdapterRSS PowerShell cmdlet.
2. Identify the processors with NUMA distance 0. These are the cores in the adapter's local
NUMA Node/Socket and will provide the best performance.
3. Adjust the RSS Base processor on each port to use a non-overlapping set of processors within
the local set of processors. You can do this manually or using the following PowerShell com-
mand:
Set-NetAdapterAdvancedProperty -Name <Adapter Name> -DisplayName
"RSS Base
Processor Number" -DisplayValue <RSS Base Proc Value>
4. Use the Get-NetAdpaterAdvancedproperty cmdlet to check that the right values have been set:
Get-NetAdpaterAdvancedproperty -Name <Adapter Name>
For Example: For a 4 port adapter with Local processors 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22,
24, 26, 28, 30, and 'Max RSS processor' of 8, set the RSS base processors to 0, 8, 16 and 24.
Optimized for quick response and low latency
lMinimize or disable Interrupt Moderation Rate.
lDisable Offload TCP Segmentation.
lDisable Jumbo Packets.
lIncrease Transmit Descriptors.
lIncrease Receive Descriptors.
lIncrease RSS Queues.
Optimized for throughput
lEnable Jumbo Packets.
lIncrease Transmit Descriptors.
lIncrease Receive Descriptors.
lOn systems that support NUMA, set the Preferred NUMA Node on each adapter to achieve better scal-
ing across NUMA nodes.

Optimized for CPU utilization
lMaximize Interrupt Moderation Rate.
lKeep the default setting for the number of Receive Descriptors; avoid setting large numbers of Receive
Descriptors.
lDecrease RSS Queues.
lIn Hyper-V environments, decrease the Max number of RSS CPUs.
Remote Storage
The remote storage features allow you to access a SAN or other networked storage using Ethernet protocols.
This includes Data Center Bridging (DCB), iSCSI over DCB, and Fibre Channel over Ethernet (FCoE).
DCB (Data Center Bridging)
Data Center Bridging (DCB) is a collection of standards-based extensions to classical Ethernet. It provides a
lossless data center transport layer that enables the convergence of LANs and SANs onto a single unified
fabric.
Furthermore, DCB is a configuration Quality of Service implementation in hardware. It uses the VLAN priority
tag (802.1p) to filter traffic. That means that there are 8 different priorities that traffic can be filtered into. It also
enables priority flow control (802.1Qbb) which can limit or eliminate the number of dropped packets during
network stress. Bandwidth can be allocated to each of these priorities, which is enforced at the hardware level
(802.1Qaz).
Adapter firmware implements LLDP and DCBX protocol agents as per 802.1AB and 802.1Qaz respectively.
The firmware based DCBX agent runs in willing mode only and can accept settings from a DCBX capable
peer. Software configuration of DCBX parameters via dcbtool/lldptool are not supported.
iSCSI Over DCB
Intel® Ethernet adapters support iSCSI software initiators that are native to the underlying operating system.
In the case of Windows, the Microsoft iSCSI Software Initiator, enables connection of a Windows host to an
external iSCSI storage array using an Intel Ethernet adapter.
In the case of Open Source distributions, virtually all distributions include support for an Open iSCSI Software
Initiator and Intel® Ethernet adapters will support them. Please consult your distribution documentation for
additional configuration details on their particular Open iSCSI initiator.
Intel® 82599 and X540-based adapters support iSCSI within a Data Center Bridging cloud. Used in
conjunction with switches and targets that support the iSCSI/DCB application TLV, this solution can provide
guaranteed minimum bandwidth for iSCSI traffic between the host and target. This solution enables storage
administrators to segment iSCSI traffic from LAN traffic, similar to how they can currently segment FCoE
from LAN traffic. Previously, iSCSI traffic within a DCB supported environment was treated as LAN traffic by
switch vendors. Please consult your switch and target vendors to ensure that they support the iSCSI/DCB
application TLV.

Intel®Ethernet FCoE (Fibre Channel over Ethernet)
Fibre Channel over Ethernet (FCoE) is the encapsulation of standard Fibre Channel (FC) protocol frames as
data within standard Ethernet frames. This link-level encapsulation, teamed with an FCoE-aware Ethernet-to-
FC gateway, acts to extend an FC fabric to include Ethernet-based host connectivity. The FCoE specification
focuses on encapsulation of FC frames specific to storage class traffic, as defined by the Fibre Channel FC-4
FCP specification.
NOTE: Support for new operating systems will not be added to FCoE. The last operating system
versions that support FCoE are as follows:
lMicrosoft*Windows Server* 2012 R2
lRHEL 7.2
lRHEL 6.7
lSLES 12 SP1
lSLES 11 SP4
lVMware* ESX 6.0
Jumbo Frames
The base driver supports FCoE mini-Jumbo Frames (2.5k bytes) independent of the LAN Jumbo Frames
setting.
FCoE VN to VN (VN2VN) Support
FCoE VN to VN, also called VN2VN, is a standard for connecting two end-nodes (ENodes) directly using
FCoE. An ENode can create a VN2VN virtual link with another remote ENode by not connecting to FC or
FCoE switches (FCFs) in between, so neither port zoning nor advance fibre channel services is required. The
storage software controls access to, and security of, LUNs using LUN masking. The VN2VN fabric may have
a lossless Ethernet switch between the ENodes. This allows multiple ENodes to participate in creating more
than one VN2VN virtual link in the VN2VN fabric. VN2VN has two operational modes: Point to Point (PT2PT)
and Multipoint.
NOTE: The mode of operation is used only during initialization.

Point to Point (PT2PT) Mode
In Point to Point mode, there are only two ENodes, and they are connected either directly or through a
lossless Ethernet switch: