Bug 690910 - opensm.conf guid entry must be incremented by 1 from the actual value to work correctly
Summary: opensm.conf guid entry must be incremented by 1 from the actual value to work...
Keywords:
Status: CLOSED NOTABUG
Alias: None
Product: Red Hat Enterprise Linux 6
Classification: Red Hat
Component: opensm
Version: 6.0
Hardware: x86_64
OS: Unspecified
unspecified
unspecified
Target Milestone: rc
: ---
Assignee: Doug Ledford
QA Contact: Red Hat Kernel QE team
URL:
Whiteboard:
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2011-03-25 18:52 UTC by Kim van der Riet
Modified: 2011-03-25 20:05 UTC (History)
0 users

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Environment:
Last Closed: 2011-03-25 20:05:07 UTC
Target Upstream Version:


Attachments (Terms of Use)

Description Kim van der Riet 2011-03-25 18:52:34 UTC
When configuring opensm to work on RHEL 6.0, I found that the guid value that must be set in the opensm.conf file must be a value of 1 higher than the actual guid to correctly start.

On my RHEL-6 x86_64 box:

# ibv_devices 
    device          	   node GUID
    ------          	----------------
    mlx4_0          	0002c90300076cb2

# cat /sys/class/infiniband/mlx4_0/node_guid 
0002:c903:0007:6cb2

However if I use this value in the opensm.conf file:

<snip>
# The port GUID on which the OpenSM is running
guid 0x0002c90300076cb2
</snip>

I get the following error:
 [1ECF4700] 0x01 -> osm_vendor_open_port: ERR 5422: Unable to find requested CA guid 0x2c90300076cb2

However, if I set the conf file to:
<snip>
# The port GUID on which the OpenSM is running
guid 0x0002c90300076cb3
</snip>

then the service starts correctly.

Package: opensm-3.3.5-1.el6.x86_64

Comment 2 RHEL Program Management 2011-03-25 19:17:49 UTC
This request was evaluated by Red Hat Product Management for
inclusion in the current release of Red Hat Enterprise Linux.
Because the affected component is not scheduled to be updated
in the current release, Red Hat is unfortunately unable to
address this request at this time. Red Hat invites you to
ask your support representative to propose this request, if
appropriate and relevant, in the next release of Red Hat
Enterprise Linux. If you would like it considered as an
exception in the current release, please ask your support
representative.

Comment 3 Doug Ledford 2011-03-25 20:05:07 UTC
The node_guid is not the same as the port guid.  You need the guid for the port you want opensm to attach to.  If you look at ibv_devinfo you will see something like this:

hca_id:	mlx4_0
	transport:			InfiniBand (0)
	fw_ver:				2.6.648
	node_guid:			0002:c903:0004:8ca0
	sys_image_guid:			0002:c903:0004:8ca3

The node guid is a base number, the first port on the card is one higher than the node guid, the second port on the card is two higher than the node guid, and then the sys image guid is one higher still yet.


Note You need to log in before you can comment on or make changes to this bug.