Bryan wrote:
On Sun, Oct 25, 2009 at 10:22 AM, Vladimir Sokolovsky
<[email protected]> wrote:
Woodruff, Robert J wrote:
Sending this to the EWG openfabrics list,
since this seems to be an OFED build/installation issue
rather than a general code problem.

One thing that you might try is to instead of copying the
entire build directory and re-runing ./install.pl -c ofed.conf
on each system, instead, after building on one node,
just copy the binrary RPMS directory and the uninstall script to the other
nodes,
Then just run the uninstall script and
install the RPMS manually... e.g,
./uninstall.sh
cd ./RPMS/redhat-release-xxxx/x86_64
rpm -i *

This method has worked for me in the past.
woody


-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of Bryan
Sent: Thursday, October 22, 2009 11:22 AM
To: [email protected]
Subject: OFED 1.4.2 requires rebuild of kernel modules on each node

I was referred to this list by the general mailing list on OFED.
Emailing from my personal address since Lotus Notes insists that
anything it sends has to contain some portion of HTML.

This problem was observed on Red Hat Enterprise Linux 5 update 3.  I
searched the list but did not see anything immediately applicable.
We've seen issues similar in the past where we were able to modify the
script to solve an RPM that didn't match the expected naming scheme,
but did not see anything immediately when looking at the scripts for
this version.

Copied from an internal bug reporting tool:

On installing OFED 1.4.2, the tarball was extracted, in directory the code
was extracted to, ./install.pl was run and all components of OFED were
build/installed with the default settings.  Then this directory was
copied to another node, and ./install.pl -c ofed.conf was run.  Previously
this would just do the install of the already built components, but with
OFED 1.4.2, the kernel RPM gets re-built when this is done.

This means that the build tools have to be on each node, and that
deployment of OFED takes longer.

Bryan Reese -- [email protected]
e1350 Linux Cluster Test Engineer
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Hi Bryan,
This can happen when some of the selected kernel components (in ofed.conf)
are not supported on the current kernel.
Have you seen the following message while running install.pl script?
"<package> is not available on this platform"

Send me please your ofed.conf and the kernel version.

Regards,
Vladimir



Hi Vladimir,

Sorry if a couple people are getting this twice, I'm not used to using
gmail with mailing lists.

We're running Red Hat Enterprise Linux 5 Update 3 with the x86_64 SMP
kernel 2.6.18-128.el5.

The below is what we see.  There is no error.  It goes on like this
afterwards with everything else getting installed with no build.

Build ofa_kernel RPM
Running rpmbuild --rebuild  --define '_topdir /var/tmp/OFED_topdir'
--define 'configure_options   --with-core-mod --with-user_mad-mod
--with-user_access-mod --with-addr_trans-mod --with-mthca-mod
--with-mlx4-mod --with-mlx4_en-mod --with-cxgb3-mod --with-nes-mod
--with-ipath_inf-mod --with-ipoib-mod --with-sdp-mod --w
ith-srp-mod --with-srp-target-mod --with-rds-mod --with-qlgc_vnic-mod
--with-iser-mod --with-nfsrdma-mod' --define 'build_kernel_ib 1'
--define 'build_kernel_ib_devel 1' --define 'KVERSION 2.6.18-128.el5'
--define 'K_SRC /lib/modules/2.6.18-128.el5/build' --define
'network_dir /etc/sysconfig/network-scripts' --define
'_prefix /usr' --define '__arch_install_post %{nil}'
/cluster/software/OFED/OFED-1.4.2/SRPMS/ofa_kernel-1.4.2-ofed1.4.2.src.rpm
Install kernel-ib RPM:
Running rpm -iv
/cluster/software/OFED/OFED-1.4.2/RPMS/redhat-release-5Server-5.3.0.3/x86_64/kernel-ib-1.4.2-2.6.18_128.el5.x86_64.rpm

kernel-ib=y
kernel-ib-devel=y
ib-bonding=y
ib-bonding-debuginfo=y
libibverbs=y
libibverbs-devel=y
libibverbs-devel-static=y
libibverbs-utils=y
libibverbs-debuginfo=y
libmthca=y
libmthca-devel-static=y
libmthca-debuginfo=y
libmlx4=y
libmlx4-devel=y
libmlx4-debuginfo=y
libcxgb3=y
libcxgb3-devel=y
libcxgb3-debuginfo=y
libnes=y
libnes-devel-static=y
libnes-debuginfo=y
libipathverbs=y
libipathverbs-devel=y
libipathverbs-debuginfo=y
libibcm=y
libibcm-devel=y
libibcm-debuginfo=y
libibcommon=y
libibcommon-devel=y
libibcommon-static=y
libibcommon-debuginfo=y
libibumad=y
libibumad-devel=y
libibumad-static=y
libibumad-debuginfo=y
libibmad=y
libibmad-devel=y
libibmad-static=y
libibmad-debuginfo=y
ibsim=y
ibsim-debuginfo=y
librdmacm=y
librdmacm-utils=y
librdmacm-devel=y
librdmacm-debuginfo=y
libsdp=y
libsdp-devel=y
libsdp-debuginfo=y
opensm=y
opensm-libs=y
opensm-devel=y
opensm-debuginfo=y
opensm-static=y
compat-dapl=y
compat-dapl-devel=y
dapl=y
dapl-devel=y
dapl-devel-static=y
dapl-utils=y
dapl-debuginfo=y
perftest=y
mstflint=y
tvflash=y
qlvnictools=y
sdpnetstat=y
srptools=y
rds-tools=y
rnfs-utils=y
ibutils=y
infiniband-diags=y
qperf=y
qperf-debuginfo=y
ofed-docs=y
ofed-scripts=y
tgt-generic=y
mpi-selector=y
mvapich_gcc=y
mvapich2_gcc=y
openmpi_gcc=y
mpitests_mvapich_gcc=y
mpitests_mvapich2_gcc=y
mpitests_openmpi_gcc=y
ibvexdmtools=y
qlgc_vnic_daemon=y
core=y
mthca=y
mlx4=y
mlx4_en=y
cxgb3=y
nes=y
ipath=y
ipoib=y
sdp=y
srp=y
srpt=y
rds=y
qlgc_vnic=y
iser=y
nfsrdma=y
mvapich2_conf_impl=ofa
mvapich2_conf_romio=1
mvapich2_conf_shared_libs=1
mvapich2_conf_ckpt=0
mvapich2_conf_vcluster=small


Hi Bryan,
Please try the following patch to the install.pl script:

diff --git a/install.pl b/install.pl
index d12733a..56ceaef 100755
--- a/install.pl
+++ b/install.pl
@@ -2452,6 +2452,10 @@ sub module_in_rpm
         return 1;
     }

+    if ($module eq "nfsrdma") {
+        $module = "xprtrdma";
+    }
+
open(LIST, "rpm -qlp $package |") or die "Can't run 'rpm -qlp $package': $!\n";
     while (<LIST>) {
         if (/$module[a-z_]*.ko/) {

Regards,
Vladimir
_______________________________________________
ewg mailing list
[email protected]
http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ewg

Reply via email to