hi all

i have a genuine Mellanox SX6012 Switch that i bought on ebay.

I have the serial #, manufacture data is 2014

There is no issue with the switch it works fine!

i have sent email to support@mellanox.com and the online form several times to ask for access to download the updated software

yet mellanox does not respond.

DOES ANY ONE HAVE ACCESS? AND CAN DOWNLOAD MLNX-OS for a SX6012 switch for me and upload it somewhere?

Furrious with Mellanox!  their support doesnt exist.  contract or otherwise!

New Values              Current Values

        Node  GUID: 0002c90300ad6280 0002c90300ad6280

        Port1 GUID:     0000000000000000        0000000000000000

        Port2 GUID:     0000000000000000        0000000000000000

        Sys.Image GUID: 0002c90300ad6280 0002c90300ad6280

        Port1 MAC:          0002c9ad6280            0002c9ad6280

        Port2 MAC:          002c9ad62e0b            0002c9ad62e0

Image type: FS2

FW Version:      9.2.6000

Device ID: 51000

Description: Node             Sys image

GUIDs:           0002c90300ad6280 0002c90300ad6280

Description: Base             Switch

MACs: 0002c9ad6280     0002c9ad62e0

VSD: n/a

PSID:            MT_1260110026

hi

do you have access to mellanox to get the MLNX-OS?

if so, i can show you how to update the firmware of your switch back to mellanox, by doing a custom firmware build.  i  have done it sucessfully on 4 x SX6012 and 6005 Switches.  But need the MLNX-OS software to install on the switch, and mellanox kindly wont give me access to download MLNX-OS!

Angry!

hi

did you get the image to download off the sx6018?  if so can i have a copy of the MLNX-OS software please?

i got an errors when trying to restart openibd after install.

centos 7.1

rpm -qa | grep rdma

librdmacm-utils-1.0.21mlnx-OFED.3.0.1.5.2.x86_64

librdmacm-devel-1.0.21mlnx-OFED.3.0.1.5.2.x86_64

librdmacm-1.0.21mlnx-OFED.3.0.1.5.2.x86_64

restart log:

Dec  8 15:17:55 srv root[24091]: openibd: running in manual mode

Dec  8 15:17:56 srv kernel: Compat-mlnx-ofed backport release: 9032737

Dec  8 15:17:56 srv kernel: Backport based on mlnx_ofed/mlnx_rdma.git 9032737

Dec  8 15:17:56 srv kernel: compat.git: mlnx_ofed/mlnx_rdma.git

Dec  8 15:17:56 srv kernel: mlx4_core: Mellanox ConnectX core driver v3.1-1.0.3 (29 Sep 2015)

Dec  8 15:17:56 srv kernel: mlx4_core: Initializing 0000:02:00.0

Dec  8 15:18:00 srv kernel: mlx4_core: device is working in RoCE mode: Roce V1

Dec  8 15:18:00 srv kernel: mlx4_core: gid_type 1 for UD QPs is not supported by the devicegid_type 0 was chosen instead

Dec  8 15:18:00 srv kernel: mlx4_core: UD QP Gid type is: V1

Dec  8 15:18:02 srv kernel: mlx4_core 0000:02:00.0: PCIe link speed is 8.0GT/s, device supports 8.0GT/s

Dec  8 15:18:02 srv kernel: mlx4_core 0000:02:00.0: PCIe link width is x8, device supports x8

Dec  8 15:18:02 srv kernel: mlx4_en: Mellanox ConnectX HCA Ethernet driver v3.1-1.0.3 (29 Sep 2015)

Dec  8 15:18:02 srv kernel: mlx4_en 0000:02:00.0: registered PHC clock

Dec  8 15:18:02 srv kernel: mlx4_en 0000:02:00.0: Activating port:1

Dec  8 15:18:02 srv kernel: mlx4_en: 0000:02:00.0: Port 1: Using 48 TX rings

Dec  8 15:18:02 srv kernel: mlx4_en: 0000:02:00.0: Port 1: Using 4 RX rings

Dec  8 15:18:02 srv kernel: mlx4_en: 0000:02:00.0: Port 1:   frag:0 - size:1522 prefix:0 stride:1536

Dec  8 15:18:02 srv kernel: mlx4_en: 0000:02:00.0: Port 1: Initializing port

Dec  8 15:18:02 srv kernel: <mlx4_ib> mlx4_ib_add: mlx4_ib: Mellanox ConnectX InfiniBand driver v3.1-1.0.3 (29 Sep 2015)

Dec  8 15:18:02 srv kernel: mlx4_core 0000:02:00.0: mlx4_ib_add: allocated counter index 1 for port 1

Dec  8 15:18:02 srv systemd-udevd[24152]: renamed network interface eth2 to mlnx0

Dec  8 15:18:02 srv logger[24198]: mlnx_interface_mgr: Setting up Mellanox network interface: mlnx0

Dec  8 15:18:02 srv logger[24203]: mlnx_interface_mgr: Bringing up interface mlnx0: PASSED

Dec  8 15:18:02 srv kernel: mlx4_en: mlnx0:   frag:0 - size:1522 prefix:0 stride:1536

Dec  8 15:18:02 srv root[24326]: openibd: ERROR: Module rdma_cm belong to kernel which is not a part of MLNX_OFED, skipping...

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_send_cm_sidr_req

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_send_cm_sidr_req (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol iw_create_cm_id

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol iw_create_cm_id (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_unregister_client

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_unregister_client (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_send_cm_sidr_rep

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_send_cm_sidr_rep (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ibnl_put_msg

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ibnl_put_msg (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol rdma_translate_ip

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol rdma_translate_ip (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_cm_listen

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_cm_listen (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ibnl_add_client

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ibnl_add_client (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_destroy_cm_id

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_destroy_cm_id (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_find_cached_pkey

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_find_cached_pkey (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol iw_cm_connect

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol iw_cm_connect (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol rdma_addr_cancel

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol rdma_addr_cancel (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_cm_notify

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_cm_notify (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_detach_mcast

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_detach_mcast (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_sa_get_mcmember_rec

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_sa_get_mcmember_rec (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_register_client

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_register_client (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_sa_path_rec_get

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_sa_path_rec_get (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_create_cm_id

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_create_cm_id (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_init_ah_from_mcmember

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_init_ah_from_mcmember (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_set_client_data

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_set_client_data (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_get_cached_pkey

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_get_cached_pkey (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol iw_destroy_cm_id

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol iw_destroy_cm_id (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_query_port

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_query_port (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_send_cm_rep

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_send_cm_rep (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_get_client_data

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_get_client_data (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ibnl_put_attr

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ibnl_put_attr (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol iw_cm_disconnect

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol iw_cm_disconnect (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_query_device

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_query_device (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_cm_init_qp_attr

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_cm_init_qp_attr (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_send_cm_drep

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_send_cm_drep (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_create_qp

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_create_qp (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_sa_free_multicast

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_sa_free_multicast (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_send_cm_mra

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_send_cm_mra (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_modify_qp

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_modify_qp (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_get_cached_gid

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_get_cached_gid (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol iw_cm_accept

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol iw_cm_accept (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol iw_cm_reject

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol iw_cm_reject (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_init_ah_from_path

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_init_ah_from_path (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_destroy_qp

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_destroy_qp (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol iw_cm_init_qp_attr

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol iw_cm_init_qp_attr (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_send_cm_rtu

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_send_cm_rtu (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_send_cm_dreq

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_send_cm_dreq (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_send_cm_req

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_send_cm_req (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_sa_join_multicast

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_sa_join_multicast (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol rdma_port_get_link_layer

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol rdma_port_get_link_layer (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_query_gid

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_query_gid (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_attach_mcast

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_attach_mcast (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_find_cached_gid

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_find_cached_gid (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol iw_cm_listen

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol iw_cm_listen (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol ib_send_cm_rej

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol ib_send_cm_rej (err -22)

Dec  8 15:18:02 srv kernel: rdma_cm: disagrees about version of symbol rdma_resolve_ip

Dec  8 15:18:02 srv kernel: rdma_cm: Unknown symbol rdma_resolve_ip (err -22)

Dec  8 15:18:04 srv kernel: mlx4_en: mlnx0: Link Up

Dec  8 15:18:22 srv root[24386]: openibd: Set node_desc for mlx4_0: srv HCA-1

Hi Bruce,

What is the switch serial number? it's available from the "show inventory " cli command (chassis).

regarding no response from Mellanox, that means that you even didn't receive any initial email with a case number or something of this sort?

hm. when I add kernel support, then reinstall - and it works.

Looks like the best way to set the MAC for for the VF is  'ip link set vf <num> mac <mac>'.

hello

i submitted my request to mellanox several times (4 + now) over the last 2 weeks, have sent them 5 emails, now they bounce back from support@mellanox.com as blacklisted.

no response at all from support, no case # no response.

I even found the $595 support package for warranty, and lodged a query about buying support, not a peep.   i thought my email wasnst working, but i see yours, and many other messages. ive checked spam, not there either.

i have 1 x SX6012 which needs updating. (cant believe i paid $12K for it)  if i cant ill just throw the fkn thing in the bin after smashing it with a sledge hammer, and forget about mellanox, and infiniband forever.  im over their arrogant treatment of their customers, no wonder its a shit view of mellanox in the community.  everyone i talk to in the IT world (IBM, Fujitsu, HP) senior systems engineers and the like, have a very dimm view of infiniband. so much so they will happily rip it out of an in infrastructure when they see it.

When i was a senior system engineer at Fujitsu, (Mellanox Japan,)they jumped right up my ass when i had problems with a clients (Fuckin huge gov dep) infiniband.  But if you a lil player they dont want to know.  What an attitude.

MT1334X03746

MT1321X01949

over this mess.

eddie.notz <community@mellanox.com> wrote ..

eddie.notz

replied to the discussion

"Firmware/OS MLNX-OS IMAGE"

 

To view the discussion, visit: https://community.mellanox.com/message/6159?et=watches.email.thread#6159

>

Hi Mellanox RDMA community,

I've come across a (painfully intermittent) problem establishing RDMA_CM connections between processes running on linux machines.

The application involves 8 servers and 40 clients. Each of the 8 servers listens for 40 connections from the clients, and each client opens connections to the 8 servers. After connection setup the clients perform RDMA writes to the servers.

The servers perform the following operations (in 40 posix threads) during connection setup:

  rdma_create_event_channel ()

  rdma_create_id ()

  rdma_bind_addr ()

  rdma_listen ()

  rdma_get_cm_event ()  [receives RDMA_CM_EVENT_CONNECT_REQUEST]

  rdma_ack_cm_event ()

  rdma_create_qp ()

  setup PDs, CQs, etc

  rdma_accept ()

  rdma_get_cm_event () [receives RDMA_CM_EVENT_ESTABLISHED]

and the clients perform the following (in 8 separate posix threads):

  rdma_create_event_channel ()

  rdma_create_id ()

  rdma_resolve_addr ()

  rdma_get_cm_event () [receives RDMA_CM_EVENT_ADDR_RESOLVED]

  rdma_ack_cm_event ()

  rdma_resolve_route ()

  rdma_get_cm_event () [receives RDMA_CM_EVENT_ROUTE_RESOLVED]

  rdma_ack_cm_event ()

  rdma_create_qp ()

  setup PDs, CQs, etc

  rdma_connect ()

  rdma_get_cm_event () [receives RDMA_CM_EVENT_ESTABLISHED]

In the vast majority of these trials this works fine, but very occasionally, a client receives a RDMA_CM_EVENT_CONNECT_ERROR event after calling rdma_connect() and the server receives a corresponding RDMA_CM_EVENT_REJECTED event after the rdma_accept() call.

I thought that the rdma library would be thread safe, but now I'm not so sure. I've not been able to gleam any further information about why the connection setup is occasionally failing other than event->status field on the server being 28.

I'd appreciate any tips on how to proceed in debugging this connection problem.

The hardware in use are Mellanox ConnectX-3 FDR10 cards connected to an SwitchX FDR 10 with the Subnet Manager running on a server. Software stack is MLNX_OFED_LINUX-3.0-2.0.1-rhel6.6-x86_64.

Thanks!

Andrew

libvma can use only ip_add_membership, but some exchanges giving only SSM.

any ways to get round this problem?

Could you try the latest 3.1-1.0.3 from http://www.mellanox.com/page/products_dyn?product_family=26&mtag=linux_sw_drivers ?

I recommend calling Mellanox to get a case opened properly and discuss support options.

In the US, the support numbers on the website are these:

Mellanox Call Center
+1 408.916.0055

Toll-free (USA only)
86-Mellanox (8663552669)

The support@mellanox.com is a valid e-mail address, but it sounds like your e-mails was filtered somehow.
Sending a request to that e-mail address is supposed to open a case - which someone responds to (probably initially via e-mail).
You should also get auto-replies, providing a case #, etc.

If you don't have a contract, the process is for Mellanox to contact you regarding support options.
It seems the e-mails replies from Mellanox are just getting filtered for some reason - maybe in your junk mail box.

This basically comes down to the implementation on how LOS is detected on the RJ45 versus the SFP+ module.

In the optical case the LOS can be detected either via the LOS or the LOS of the pin detector, in the RJ45 case LOS can only be detected via the LOS pin which is not connected hence why the the OS does see the link go down when the physical link goes down between the switch and the NIC.

If you had a GbE SFP optical module the behavior would be the same as the 10GbE SFP+ module.

does anyone know how to programm the .img file of mlnx-os to the flash (JFFS2) 1GB flash drive built into the SX10XX series of switches

ive a corrupt boot image. (thus no mlnx-os running)  can only interrupt boot process to goto u-boot

i finally have the img file from mellanox to upload.  It is .img file

Ive setup a tftp server as per this post Put JFFS2 Image to Flash - Texas Instruments Wiki  and RootFileSystemOnAJFFS2FileSystem < DULG < DENX

ive run envreset to reset the environment settings

imls shows 2 flash images

Legacy image at FF400000

Legacy image at FF600000

checksum...OK

i then try to upload via tftp FF400000 imagexxx.img (from ftp server)

it finds the image and begins to load it

=> tftp 100000 /tftpboot/TQM860L/jffs2.img

Using FEC ETHERNET device

TFTP from server 192.168.3.1; our IP address is 192.168.3.80

Filename '/tftpboot/TQM860L/jffs2.img'.

Load address: FF400000

Loading: #################################################################

  #################################################################

  #################################################################

  #################################################################

  #################################################################

  #################################################################

  ########

IT stops midway.  terminal is unresponsive.  i wait for 5 mins, as when trying to upload it via NFS or tftp results in a error message saying img in wrong format.

upon reboot

image at FF400000

Verifying checksum ... bad data CRC

when i try to run, it doesnt run, it shoudl show something liek this

# flpart

Nr | Name | Start | Size | Type | FS | Flags

---------------------------------------------------------------------

0 | U-Boot | 0 |768 KiB |U-Boot | | fixed

1 | NVRAM | 768 KiB |512 KiB |NVRAM | | fixed

2 | Kernel | 1280 KiB | 3 MiB |Linux-Kernel | |

3 | RootFS-JFFS2 | 4352 KiB | 16 MiB |Filesystem | JFFS2 | rootfs

4 | User-JFFS2 |20736 KiB | 11 MiB | Filesystem | JFFS2 |

Commands:

a) Append partition

d) Delete partition

m) Modify partition

p) Print partition table

r) Reset partition table

q) Quit

Cmd (? for help

Ive also tried NFS and tftp to upload and auto program but it says corrupt image error also.?

iget an error

bytes transfered 298732816 (11ce4d10) hex

automatic boot format for tftpboot command

ERROR: cant get kernel image

Am i to assume i have to also build a custom kernel and upload it also?

also tried

=>setenv flash_jffs2

=>run mlxlinux

wront image format foor bootm command

ERROR: cant get kernel image

so im guessing the kernel image partition is the one which is corrupt.

output from flint

flint -d  /dev/mst/SW_MT51000_0x2c90300ad6280_lid-0x0002 v

     FS2 failsafe image. Start address: 0x200000. Chunk size 0x200000:

     NOTE: The addresses below are contiguous logical addresses. Physical addresses on

           flash may be different, based on the image start address and chunk size

     /0x00000038-0x000013bb (0x001384)/ (BOOT2) - OK

     /0x000013bc-0x00002913 (0x001558)/ (BOOT2) - OK

     /0x00002914-0x00002a0b (0x0000f8)/ (Configuration) - OK

     /0x00002a0c-0x00002a4f (0x000044)/ (GUID) - OK

     /0x00002a50-0x00002ba7 (0x000158)/ (Image Info) - OK

     /0x00002ba8-0x00003b6f (0x000fc8)/ (DDR) - OK

     /0x00003b70-0x001708c3 (0x16cd54)/ (DDR) - OK

     /0x001708c4-0x001712ff (0x000a3c)/ (DDR) - OK

     /0x00171300-0x0017ca2b (0x00b72c)/ (Configuration) - OK

     /0x0017ca2c-0x0017ca57 (0x00002c)/ (Jump addresses) - OK

     /0x0017ca58-0x0017efe7 (0x002590)/ (FW Configuration) - OK

     /0x00000000-0x0017efe7 (0x17efe8)/ (Full Image) - OK

-I- FW image verification succeeded. Image is bootable.

how to fix that?

Any ideas?

I'm looking for instructions to build the latest openvswitch w/ dpdk while using connectx-3 nics from mellanox and if possible in SR-IOV mode ?

Has anyone done this?

I have problem like this but cant have access to switch cli. Can I get switch serial number from ofed or ibutils ?

iperf gives me max of 18Gb/s with th nic's direct attached

and even worse 8Gb/s with the sx1012 interconnect

hw: z800 LnkSta 5GT/s x 8       asus z10pe-d8 intel xeon server LnkSta 8GT/s x 8

os: 2 x centos6.7

can someone help an ethernet basic skilled user

Hi,

I am trying to set up IB

I have several MHGH28-XTC HCAs

1 appears to work but others do not.

The working HCA has rev A3 on the label the others show A4.

Running CentOS on the server with OFED.

I can see the adapters when I run lspci they all show:

[root@localhost ~]# lspci | grep Mel

01:00.0 InfiniBand: Mellanox Technologies MT25418 [ConnectX VPI PCIe 2.0 2.5GT/s - IB DDR / 10GigE] (rev a0)

But when I run ib_devinfo I get:

[root@localhost ~]# ibv_devinfo

No IB devices found

The working device shows:

[root@localhost ~]# ibv_devinfo

hca_id: mlx4_0

        transport:                      InfiniBand (0)

        fw_ver:                         2.7.000

        node_guid:                      0002:c903:0000:d25c

        sys_image_guid:                 0002:c903:0000:d25f

        vendor_id:                      0x02c9

        vendor_part_id:                 25418

        hw_ver:                         0xA0

        board_id:                       MT_04A0120002

        phys_port_cnt:                  2

                port:   1

                        state:                  PORT_ACTIVE (4)

                        max_mtu:                4096 (5)

                        active_mtu:             2048 (4)

                        sm_lid:                 1

                        port_lid:               2

                        port_lmc:               0x00

                        link_layer:             InfiniBand

                port:   2
                        state:                  PORT_ACTIVE (4)
                        max_mtu:                4096 (5)
                        active_mtu:             2048 (4)
                        sm_lid:                 1
                        port_lid:               3
                        port_lmc:               0x00
                        link_layer:             InfiniBand

I was hoping I could change the firmware to get these other cards working but they don't even appear?!?

Thanks for any help.

Premature posting I'm afraid.

Appear to have sorted this out (I'm getting link lights now anyway).

I used mstflint to query the PCI ID and then burn an older firmware on, 2.9.1 isn't working 2.8.6 is.