Hi Kyle,

As a temporary work around, I hard coded it to use the 2nd one on that
particular server.

Let me know when you are ready for me to test.

Thanks,

Mi
On Wed, 2012-03-14 at 10:26 -0500, Kyle Schochenmaier wrote:
> Hi Mi -
> 
> I took another look at the code here, I may have to rewrite the entire
> connection initialization routine for openib to be able to support
> this.
> Since it isnt trivial it may take me a while to pull things together.
> 
> Also, there doesn't appear to be any workarounds to using the 2nd
> device other than for me to suggest plugging in the first device,
> until this gets rewritten :-(
> 
> Unfortunately I do not have the hardware to test any of the changes
> I'll make, so if you're willing once I get rewritten you can be the
> guinea pig for testing.
> 
> More info to come.
> 
> Kyle Schochenmaier
> 
> 
> 
> On Tue, Mar 13, 2012 at 4:01 PM, Mi Zhou <[email protected]> wrote:
> > That will work!
> >
> > Thanks,
> >
> > Mi
> > On Tue, 2012-03-13 at 15:55 -0500, Kyle Schochenmaier wrote:
> >> Hi mi, it was hard coded to pick the first card if more than one was
> >> available, this was done to keep things simple, I can add the logic to
> >> poll all active cards and send you a patch, but this won't be
> >> configurable yet. We are discussing how to best handle these
> >> situations in a configurable manner right now
> >>
> >> On Mar 13, 2012 3:45 PM, "Mi Zhou" <[email protected]> wrote:
> >>         Hi,
> >>
> >>         We are setting up pvfs2 server with orangefs-2.8.5. We have 2
> >>         IB cards
> >>         in the system, but one of them is not active. When I tried to
> >>         start up
> >>         pvfs2 server, it complains about the card that is not active,
> >>         and the
> >>         server won't start up. Here is the log
> >>
> >>          [D 03/13/2012 15:28:41] Passing ib://pvfs03:3335 as BMI
> >>         listen address.
> >>         [D 03/13/2012 15:28:41] BMI_ib_initialize: init.
> >>         [D 03/13/2012 15:28:41] openib_ib_initialize: init.
> >>         [E 03/13/2012 15:28:41] Warning: get_nic_handle: found 2 HCAs,
> >>         choosing
> >>         the first.
> >>         [E 03/13/2012 15:28:41] Error: openib_ib_initialize: port
> >>         state is
> >>         IBV_PORT_DOWN but should be ACTIVE; check subnet manager.
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /opt/orangefs/orangefs/sbin/pvfs2-server(error+0xca)
> >>         [0x482faa]
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /opt/orangefs/orangefs/sbin/pvfs2-server(openib_ib_initialize
> >>         +0x3d2) [0x4841d2]
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /opt/orangefs/orangefs/sbin/pvfs2-server() [0x482550]
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /opt/orangefs/orangefs/sbin/pvfs2-server() [0x432cfb]
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /opt/orangefs/orangefs/sbin/pvfs2-server(BMI_initialize
> >>         +0x1ef)
> >>         [0x43416f]
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /opt/orangefs/orangefs/sbin/pvfs2-server() [0x412e15]
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /opt/orangefs/orangefs/sbin/pvfs2-server() [0x414703]
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /opt/orangefs/orangefs/sbin/pvfs2-server(main+0xdfc)
> >>         [0x4158cc]
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /lib64/libc.so.6(__libc_start_main
> >>         +0xfd) [0x36be81ec9d]
> >>         [E 03/13/2012 15:28:41]
> >>         [bt] /opt/orangefs/orangefs/sbin/pvfs2-server() [0x411ac9]
> >>
> >>         I wonder why it did not use the active card. Has anybody seen
> >>         this?
> >>
> >>         Thanks,
> >>
> >>         Mi Zhou
> >>         St. Jude Children's Research Hospital
> >>
> >>
> >>         Email Disclaimer:  www.stjude.org/emaildisclaimer
> >>
> >>
> >>         _______________________________________________
> >>         Pvfs2-users mailing list
> >>         [email protected]
> >>         http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users
> > --
> >
> > Mi Zhou
> > System Integration Engineer
> > Information Sciences
> > St. Jude Children's Research Hospital
> > 262 Danny Thomas Pl. MS 312
> > Memphis, TN 38105
> > 901.595.5771
> >
> >
> 
-- 

Mi Zhou
System Integration Engineer
Information Sciences
St. Jude Children's Research Hospital
262 Danny Thomas Pl. MS 312 
Memphis, TN 38105
901.595.5771


_______________________________________________
Pvfs2-users mailing list
[email protected]
http://www.beowulf-underground.org/mailman/listinfo/pvfs2-users

Reply via email to