Intel X710 (X710DA2) loop issue?

[idata]

Hi,

We order an Intel X710DA2 from http://www.provantage.com/intel-x710da2~7ITEN0KM.htm Provantage and SFPs from http://www.amazon.com/Intel-E10GSFPSR-Ethernet-Optic-10GBase-SR/dp/B002OHV01K Amazon.

The NIC and accompanied SFPs work fine with full speed (10 Gbps) when connecting with other NIC Ports.

The Issue we need to ask here is when we loop the two ports of that X710DA2 together (connect the cable from port 1 to port 2 on same X710DA), the port status show "Cable Unplugged", of course the Intel IOUtil show IO not Enable:

NIC B/D/FVen-Dev MAC IO Branding string

=== ======== ========= ============ === =======================================

1) 02/00/00 8086-1572 NA Intel(R) Ethernet Converged Network Adapter X710-

2) 02/00/01 8086-1572 NA Intel(R) Ethernet Controller X710 for 10GbE SFP+

Also, DPDK (1.8) testcmd shows only 1Gbps link connection between them.

# build/app/testpmd -c7 -n3 -- -i --nb-cores=2 --nb-ports=2

Port 0: 68:05:CA:30:51:B4

Configuring Port 1 (socket 0)

PMD: i40e_dev_tx_queue_setup(): Using full-featured tx path

PMD: i40e_dev_rx_queue_setup(): Rx Burst Bulk Alloc Preconditions are satisfied. Rx Burst Bulk Alloc function will be used on port=1, queue=0.

Port 1: 68:05:CA:30:51:B5

Checking link statuses...

Port 0 Link Up - speed 1000 Mbps - full-duplex

Port 1 Link Up - speed 1000 Mbps - full-duplex

Please explain to us about this unexpected situation.

FYI,

Our current OS is CENTOS 7 with latest controller driver from https://downloadcenter.intel.com/download/24411/Network-Adapter-Driver-for-PCI-E-40-Gigabit-Network-Connections-under-Linux- Intel® Download Center

Lspci Info:

# lspci | egrep -i --color 'network|ethernet'

02:00.0 Ethernet controller: Intel Corporation Ethernet 10G 2P X710 Adapter (rev 01)

02:00.1 Ethernet controller: Intel Corporation Ethernet 10G 2P X710 Adapter (rev 01)

Lshw Info:

# lshw -class network

*-network:0

description: Ethernet interface

product: Ethernet 10G 2P X710 Adapter

vendor: Intel Corporation

physical id: 0

bus info: pci@0000:02:00.0

logical name: enp2s0f0

version: 01

serial: 68:05:ca:30:51:b4

width: 64 bits

clock: 33MHz

capabilities: pm msi msix pciexpress vpd bus_master cap_list rom ethernet physical autonegotiation

configuration: autonegotiation=off broadcast=yes driver=i40e driverversion=0.3.30-k firmware=f4.22 a1.1 n40.25 e8000143f latency=0 link=no multicast=yes

resources: irq:17 memory:f5000000-f57fffff memory:f5808000-f580ffff memory:f7580000-f75fffff

*-network:1

description: Ethernet interface

product: Ethernet 10G 2P X710 Adapter

vendor: Intel Corporation

physical id: 0.1

bus info: pci@0000:02:00.1

logical name: enp2s0f1

version: 01

serial: 68:05:ca:30:51:b5

<p styl...

[st4]

Hi x710,

Thank you for the post, Just to double check looping the two ports of that X710DA, are you referring to connecting the cable from port 1 to port 2 of the same X710DA? This kind of connection yields limited performance ..

rgds,

wb

[idata]

Hi wb_Intel,

wb_Intel wrote:

Hi x710,

Thank you for the post, Just to double check looping the two ports of that X710DA, are you referring to connecting the cable from port 1 to port 2 of the same X710DA? This kind of connection yields limited performance ..

rgds,

wb

Yep, we connect the cable from port 1 to port 2 on same X710DA. Is there any document/note refer to this limited performance case?

Regards,

Tu

[st4]

Hi Tu,

Thank you for the clarification. I will further check on your setup here.

rgds,

wb

[st4]

Hi Tu,

Please check if you are using the latest firmware, here is the latest NVMupdate site

https://downloadcenter.intel.com/download/24769/NVM-Update-Utility-for-Intel-Ethernet-Converged-Network-Adapter-XL710-X710-Series Intel® Download Center

thanks,

wb

[idata]

Thanks.

[root@linuxsrv13 Linux_x64]# ./nvmupdate64e

Intel(R) Ethernet NVM Update Tool

NVMUpdate version 1.24.33.08

Copyright (C) 2013 - 2015 Intel Corporation.

WARNING: TO AVOID DAMAGE TO YOUR DEVICE, DO NOT EXIT OR REBOOT OR POWER OFF THE SYSTEM DURING THIS UPDATE

Inventory in progress. Please wait [.......-**]

Num Description Device-Id B:D Adapter Status

=== ====================================== ========= ===== ====================

01) Intel(R) Ethernet Converged Network Ad 8086-1572 02:00 Up to date

02) Intel(R) I210 Gigabit Network Connecti 8086-1533 05:00 Update not available

03) Intel(R) I210 Gigabit Network Connecti 8086-1533 06:00 Update not available

Tool execution completed with the following status: All operations completed successfully

[idata]

Still 1000Mbps. Issue still there.

Interactive-mode selected

Configuring Port 0 (socket 0)

PMD: i40e_dev_tx_queue_setup(): Using full-featured tx path

PMD: i40e_dev_rx_queue_setup(): Rx Burst Bulk Alloc Preconditions are satisfied. Rx Burst Bulk Alloc function will be used on port=0, queue=0.

Port 0: 68:05:CA:30:51:B4

Configuring Port 1 (socket 0)

PMD: i40e_dev_tx_queue_setup(): Using full-featured tx path

PMD: i40e_dev_rx_queue_setup(): Rx Burst Bulk Alloc Preconditions are satisfied. Rx Burst Bulk Alloc function will be used on port=1, queue=0.

Port 1: 68:05:CA:30:51:B5

Checking link statuses...

Port 0 Link Up - speed 1000 Mbps - full-duplex

Port 1 Link Up - speed 1000 Mbps - full-duplex

Done

[SYeo3]

Hi x710,

Thank you for sharing the test results with us.

I have noted this and we will further check on this issue.

Sincerely,

Sandy

[SYeo3]

Hi x710,

We would like to understand more about the network card and your system settings. Please answer details below:

1. Serial number and part number from the white sticker. (See adapter photo http://www.intel.com/support/network/adapter/pro100/sb/CS-031543.htm here)

2. Is the CentOS running on a bare metal system, or in a virtual machine.

If in a VM, what the Virtual Host (ie VMWare).

3. How long is the fiber cable?

4. What is the current version of the NVM on the adapter?

We look forward to your reply.

Sincerely,

Sandy

[idata]

Thanks for your replying.

FYI,

1.

2. CentOS 7 is running on a bare metal system.

3. Fiber cable length is 5m.

4. Current version of the NVM on the adapter is 1.24.33.08

5. Firmware version showing after ethtool -i command is

firmware-version: f4.33.31377 a1.2 n4.42 e191b

Regards,

Tu

[SYeo3]

Dear Tu,

Thanks for providing the details. This helps us understand your environment.

The sticker where you can find the serial number is located at the front panel.

Format: 12-digit(MAC Address) 5-digit 6-3 digit number.

We'll wait for your reply.

Sincerely,

Sandy

[idata]

Hi all the number is :

6805CA3051B4

004 4614ad

H58362 001

Regards,

Tu

[SYeo3]

Hi Tu,

Thanks for your quick response. I'll check on this and will update you on our findings.

Sincerely,

Sandy

[SYeo3]

Hi Tu,

Upon checking, we think that its the driver problem or the driver is not loaded. Please check if you are using the latest driver.

The latest driver is available for download https://downloadcenter.intel.com/download/24411/Network-Adapter-Driver-f or-PCI-E-40-Gigabit-Network-Connections-under-Linux- here.

Hope this help resolve the issue.

Sincerely,

Sandy

[idata]

Thanks for your response,

I've installed the latest version and this issue still there .

Currently, I use the ethtool for checking driver version. Is there any tips for deeper diagnosing? Please advice me.

ethtool -i enp2s0f0

driver: i40e

version: 1.2.38

firmware-version: f4.33.31377 a1.2 n4.42 e191b

bus-info: 0000:02:00.0

supports-statistics: yes

supports-test: yes

supports-eeprom-access: yes

supports-register-dump: yes

supports-priv-flags: yes

Regards,

Tu

[st4]

Hi Tu,

Thank you for the update. Will further check on this.

rdgs,

wb

[st4]

Hi Tu,

For further diagnostics, please do the following:

ethtool ' with no additional arguments

this can tell the supported link modes and whether auto-negotiation is enabled

thanks,

wb

[idata]

FYI,

Maybe because this ETH-Dev does not support auto-neg, therefore it cannot work at the highest speed, doesn't it?

ethtool enp2s0f0

Settings for enp2s0f0:

Supported ports: [ ]

Supported link modes: 1000baseT/Full

10000baseT/Full

Supported pause frame use: Symmetric

Supports auto-negotiation: No

Advertised link modes: 1000baseT/Full

10000baseT/Full

Advertised pause frame use: No

Advertised auto-negotiation: No

Speed: Unknown!

Duplex: Unknown! (255)

Port: Other

PHYAD: 0

Transceiver: external

Auto-negotiation: off

Supports Wake-on: d

Wake-on: d

Current message level: 0x0000000f (15)

drv probe link timer

Link detected: no

[st4]

Hi Tu,

Thank you for the update. let me further check on this.

rgds,

wb

[oherm]

Hi

I am facing a similar issue with an X710 I am using at my lab, which is not syncing to speed changes in the link peer.

The ethtool shows no autonegotiation support, and when trying to activate autoneg it replies with error and there are warnings in /var/log/messages.

I am using i40e driver 1.2.38 on CentOs6.5.

[root@pe03-dut ~]# ethtool eth-tg8

Settings for eth-tg8:

Supported ports: [ FIBRE ]

Supported link modes: 1000baseT/Full

10000baseT/Full

Supported pause frame use: Symmetric

Supports auto-negotiation: No

Advertised link modes: 1000baseT/Full

10000baseT/Full

Advertised pause frame use: No

Advertised auto-negotiation: No

Speed: 10000Mb/s

Duplex: Full

Port: FIBRE

PHYAD: 0

Transceiver: external

Auto-negotiation: off

Supports Wake-on: d

Wake-on: d

Current message level: 0x0000000f (15)

drv probe link timer

[root@pe03-dut ~]# ethtool -s eth-tg8 autoneg on

Cannot set new settings: Invalid argument

not setting autoneg

[root@pe03-dut ~]# cat /var/log/messages

...

May 3 23:32:06 localhost kernel: [59453.619472] i40e 0000:83:00.0: eth-tg8: If you are not setting advertising to 1020 then you may have an old version of ethtool. Please update.

May 3 23:32:06 localhost kernel: [59453.646562] i40e 0000:83:00.0: eth-tg8: Autoneg not supported on this phy

Thanks

Ofir

[st4]

Hi Ofirh,

Thank you for the info. we are checking on this.

rgds,

wb