[Gluster-users] hanging "df" (3.1, infiniband)

Lana Deere lana.deere at gmail.com
Tue Oct 19 23:41:22 UTC 2010


I believe it's OFED but I suppose I'm not certain.  The initial
installation of these systems included IB support which was installed
by the vendor along with the OS (CentOS 5.5) and which was turned off
("chkconfig openibd off") when delivered.

I don't seem to have an "ofed_info" command.  Also, my openib.conf is
in /etc/ofed rather than /etc/infiniband.  It contains:

[root at storage0 ~]# cat /etc/ofed/openib.conf
# Load IPoIB
IPOIB_LOAD=yes
# Load SDP module
SDP_LOAD=yes
# Load the RDS protocol driver
RDS_LOAD=yes
# Load SRP module
SRP_LOAD=yes
# Load iSER module
ISER_LOAD=yes
# Should we modify the system mtrr registers?  We may need to do this if you
# get messages from the ib_ipath driver saying that it couldn't enable
# write combining for the PIO buffs on the card.
FIXUP_MTRR_REGS=no

.. Lana (lana.deere at gmail.com)






On Tue, Oct 19, 2010 at 7:25 PM, Joe Landman
<landman at scalableinformatics.com> wrote:
> On 10/19/2010 07:21 PM, Lana Deere wrote:
>>
>> This does exit right away, with the same failure as ibv_devinfo:
>>
>> [root at storage0 ~]# ib_send_bw
>> ------------------------------------------------------------------
>>                     Send BW Test
>> Connection type : RC
>> libibverbs: Warning: no userspace device-specific driver found for
>> /sys/class/infiniband_verbs/uverbs0
>> No IB devices found
>
> you are missing ib verbs support.  Which is strange, as its in your module
> list:
>
> ib_uverbs              68720  2 rdma_ucm,ib_ucm
>
> What does your /etc/infiniband/openib.conf look like (this is OFED, right?
>  Could you do an
>
>        ofed_info | head  -1
>
> and send output?
>
> --
> Joseph Landman, Ph.D
> Founder and CEO
> Scalable Informatics, Inc.
> email: landman at scalableinformatics.com
> web  : http://scalableinformatics.com
>       http://scalableinformatics.com/jackrabbit
> phone: +1 734 786 8423 x121
> fax  : +1 866 888 3112
> cell : +1 734 612 4615
>



More information about the Gluster-users mailing list