Can't start opensmd service

Can’t start opensmd service, the log file in ’ var/log/opensm.log ’ show as:

Aug 08 14:44:35 564382 [41B0740] 0x80 → OpenSM 5.4.0.MLNX20190516.1943e24

Entering DISCOVERING state

Aug 08 14:44:35 572847 [41B0740] 0x02 → osm_vendor_init: 1000 pending umads specified

Aug 08 14:44:35 572876 [41B0740] 0x02 → osm_vendor_init: 1000 pending umads specified

Aug 08 14:44:35 573545 [41B0740] 0x80 → Entering DISCOVERING state

Aug 08 14:44:35 573594 [41B0740] 0x02 → osm_vendor_bind: Mgmt class 0x81 binding to port GUID 0xba599ffffe4023ef

Aug 08 14:44:35 578617 [41B0740] 0x01 → osm_vendor_bind: ERR 5426: Unable to register class 129 version 1

Aug 08 14:44:35 578631 [41B0740] 0x01 → osm_sm_mad_ctrl_bind: ERR 3118: Vendor specific bind failed

Aug 08 14:44:35 578638 [41B0740] 0x01 → osm_sm_bind: ERR 2E10: SM MAD Controller bind failed (IB_ERROR)

Error from osm_opensm_bind (0x2A)

Perhaps another instance of OpenSM is already running

Aug 08 14:44:35 578646 [41B0740] 0x01 → perfmgr_mad_unbind: ERR 5405: No previous bind

Aug 08 14:44:35 578648 [41B0740] 0x01 → osm_congestion_control_shutdown: ERR C108: No previous bind

Aug 08 14:44:35 578705 [41B0740] 0x01 → osm_sa_mad_ctrl_unbind: ERR 1A11: No previous bind

Exiting SM

Aug 08 14:44:35 579084 [41B0740] 0x80 → Exiting SM

And the opensmd conf file in /etc/opensm/opensm.conf as below:

guid 0xba599ffffe4023ef

guid 0xba599ffffe4023ee

There are no demaon progress like opensmd already running background.

And ’ connectx_port_config ’ shows:

ConnectX PCI devices :

|----------------------------|

|----------------------------|

No devices found, quiting…

ibstat in the host shows:

CA ‘mlx5_0’

CA type: MT41682

Number of ports: 1

Firmware version: 18.25.6608

Hardware version: 0

Node GUID: 0xb8599f03004023ee

System image GUID: 0xb8599f03004023ee

Port 1:

State: Active

Physical state: LinkUp

Rate: 40

Base lid: 0

LMC: 0

SM lid: 0

Capability mask: 0x00010000

Port GUID: 0xba599ffffe4023ee

Link layer: Ethernet

CA ‘mlx5_1’

CA type: MT41682

Number of ports: 1

Firmware version: 18.25.6608

Hardware version: 0

Node GUID: 0xb8599f03004023ef

System image GUID: 0xb8599f03004023ee

Port 1:

State: Active

Physical state: LinkUp

Rate: 100

Base lid: 0

LMC: 0

SM lid: 0

Capability mask: 0x00010000

Port GUID: 0xba599ffffe4023ef

Link layer: Ethernet

Hello Logan,

Many thanks for posting your question on the Mellanox Community.

Based on your opensm.conf you have two GUID’s configured for binding the ‘opensm’ process to. This is not possible. You can only specify one GUID per configuration file.

if your purpose is to start two OpenSM’s on one host, you need to create a secondary opensm.conf file and start the 2nd OpenSM daemon pointing to that configuration file → opensm -F <path_to_secondary_opensm.conf_file>

Many thanks,

~Mellanox Technical Support