Hi, I''m using Centos 4.4 (EL 4 clone ) with stock rpm from lustre. trying just to initially setup an mds I have got the ost''s up,however the mds fail to initialize my lmc file is --add node --node mds02.fabrixtv.local --add net --node mds02.fabrixtv.local --nid mds02.fabrixtv.local --nettype lnet --add mds --node mds02.fabrixtv.local --mds mds-test --dev /dev/hda3 --format --fstype ldiskfs --add lov --lov lov01 --mds mds-test --add node --node ost02.fabrixtv.local --add net --node ost02.fabrixtv.local --nid ost02.fabrixtv.local --nettype lnet--add node --node ost01.fabrixtv.local --add net --node ost01.fabrixtv.local --nid ost01.fabrixtv.local --nettype lnet--add ost --node ost02.fabrixtv.local --mds mds-test --ost ost02 --lov lov01 --dev /dev/hda3 uname -a Linux mds02.fabrixtv.local 2.6.9-42.EL_lustre.1.4.7smp #1 SMP Tue Aug 22 16:29:06 MDT 2006 i686 i686 i386 GNU/Linux output of lconf -v ./config.xml configuring for host: [''mds02.fabrixtv.local''] Checking XML modification time + debugfs -c -R ''stat /LOGS'' /dev/hda3 2>&1 | grep mtime setting /proc/sys/net/core/rmem_max to at least 16777216 setting /proc/sys/net/core/wmem_max to at least 16777216 Service: network NET_mds02.fabrixtv.local_tcp T_mds02.fabrixtv.local_tcp_UUID loading module: libcfs srcdir None devdir libcfs + /sbin/modprobe libcfs loading module: lnet srcdir None devdir lnet + /sbin/modprobe lnet + /sbin/modprobe lnet loading module: ksocklnd srcdir None devdir klnds/socklnd + /sbin/modprobe ksocklnd Service: ldlm ldlm ldlm_UUID loading module: lvfs srcdir None devdir lvfs + /sbin/modprobe lvfs loading module: obdclass srcdir None devdir obdclass + /sbin/modprobe obdclass loading module: ptlrpc srcdir None devdir ptlrpc + /sbin/modprobe ptlrpc Service: mdsdev MDD_mds-test_mds02.fabrixtv.local -test_mds02.fabrixtv.local_UUID original inode_size 0 stripe_count 1 inode_size 512 loading module: mdc srcdir None devdir mdc + /sbin/modprobe mdc loading module: osc srcdir None devdir osc + /sbin/modprobe osc loading module: lov srcdir None devdir lov + /sbin/modprobe lov loading module: mds srcdir None devdir mds + /sbin/modprobe mds loading module: ldiskfs srcdir None devdir ldiskfs + /sbin/modprobe ldiskfs loading module: fsfilt_ldiskfs srcdir None devdir lvfs + /sbin/modprobe fsfilt_ldiskfs + sysctl lnet/debug_path /tmp/lustre-log-mds02.fabrixtv.local + /usr/sbin/lctl modules > /tmp/ogdb-mds02.fabrixtv.local Service: network NET_mds02.fabrixtv.local_tcp T_mds02.fabrixtv.local_tcp_UUID NETWORK: NET_mds02.fabrixtv.local_tcp T_mds02.fabrixtv.local_tcp_UUID tcp mds02.fabrixtv.local Service: ldlm ldlm ldlm_UUID Service: mdsdev MDD_mds-test_mds02.fabrixtv.local -test_mds02.fabrixtv.local_UUID original inode_size 0 stripe_count 1 inode_size 512 MDSDEV: mds-test mds-test_UUID /dev/hda3 ldiskfs 0 yes + /usr/sbin/lctl attach mdt MDT MDT_UUID quit + /usr/sbin/lctl cfg_device MDT setup quit + dumpe2fs -f -h /dev/hda3 no external journal found for /dev/hda3 MDS mount options: errors=remount-ro + /usr/sbin/lctl attach mds mds-test mds-test_UUID quit + /usr/sbin/lctl cfg_device mds-test setup /dev/hda3 ldiskfs mds-test errors=remount-ro quit + /usr/sbin/lctl ignore_errors cfg_device $mds-test cleanup detach quit MDS failed to start. Check the syslog for details. (May need to run lconf --write-conf) --add ost --node ost01.fabrixtv.local --mds mds-test --ost ost01 --lov lov01 --dev /dev/hda3 output of lconf -v --write-conf ./config.xml lconf -v --write-conf ./config.xml configuring for host: [''mds02.fabrixtv.local''] Service: network NET_mds02.fabrixtv.local_tcp T_mds02.fabrixtv.local_tcp_UUID loading module: libcfs srcdir None devdir libcfs + /sbin/modprobe libcfs loading module: lnet srcdir None devdir lnet + /sbin/modprobe lnet + /sbin/modprobe lnet loading module: ksocklnd srcdir None devdir klnds/socklnd + /sbin/modprobe ksocklnd Service: ldlm ldlm ldlm_UUID loading module: lvfs srcdir None devdir lvfs + /sbin/modprobe lvfs loading module: obdclass srcdir None devdir obdclass + /sbin/modprobe obdclass loading module: ptlrpc srcdir None devdir ptlrpc + /sbin/modprobe ptlrpc Service: mdsdev MDD_mds-test_mds02.fabrixtv.local -test_mds02.fabrixtv.local_UUID original inode_size 0 stripe_count 1 inode_size 512 loading module: mdc srcdir None devdir mdc + /sbin/modprobe mdc loading module: osc srcdir None devdir osc + /sbin/modprobe osc loading module: lov srcdir None devdir lov + /sbin/modprobe lov loading module: mds srcdir None devdir mds + /sbin/modprobe mds loading module: ldiskfs srcdir None devdir ldiskfs + /sbin/modprobe ldiskfs loading module: fsfilt_ldiskfs srcdir None devdir lvfs + /sbin/modprobe fsfilt_ldiskfs Service: mdsdev MDD_mds-test_mds02.fabrixtv.local -test_mds02.fabrixtv.local_UUID original inode_size 0 stripe_count 1 inode_size 512 MDSDEV: mds-test mds-test_UUID /dev/hda3 ldiskfs yes + /usr/sbin/lctl attach mds mds-test mds-test_UUID quit + /usr/sbin/lctl cfg_device mds-test setup /dev/hda3 ldiskfs quit + /usr/sbin/lctl ignore_errors cfg_device $mds-test cleanup detach quit + losetup /dev/loop0 + losetup /dev/loop1 + losetup /dev/loop2 + losetup /dev/loop3 + losetup /dev/loop4 + losetup /dev/loop5 + losetup /dev/loop6 + losetup /dev/loop7 changing mtime of LOGS to 1159912052 + mktemp /tmp/lustre-cmd.XXXXXXXX + debugfs -w -R "mi /LOGS" </tmp/lustre-cmd.eSSA7850 /dev/hda3 Service: mdsdev MDD_mds-test_mds02.fabrixtv.local -test_mds02.fabrixtv.local_UUID original inode_size 0 stripe_count 1 inode_size 512 unloading module: fsfilt_ldiskfs + /sbin/rmmod fsfilt_ldiskfs unloading module: ldiskfs + /sbin/rmmod ldiskfs unloading module: mds + /sbin/rmmod mds unloading module: lov + /sbin/rmmod lov unloading module: osc + /sbin/rmmod osc unloading module: mdc + /sbin/rmmod mdc Service: ldlm ldlm ldlm_UUID unloading module: ptlrpc + /sbin/rmmod ptlrpc unloading module: obdclass + /sbin/rmmod obdclass unloading module: lvfs + /sbin/rmmod lvfs Service: network NET_mds02.fabrixtv.local_tcp T_mds02.fabrixtv.local_tcp_UUID unloading module: lnet + /usr/sbin/lctl network unconfigure unloading the network + /usr/sbin/lctl network unconfigure + /sbin/rmmod ksocklnd + /sbin/rmmod lnet unloading module: libcfs + /sbin/rmmod libcfs suggestions ?
On Oct 04, 2006 00:34 +0200, Itamar Ofek wrote:> I''m using Centos 4.4 (EL 4 clone ) with stock rpm from lustre. > trying just to initially setup an mds > I have got the ost''s up,however the mds fail to initialize > > suggestions ? > > MDS failed to start. Check the syslog for details.^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Cheers, Andreas -- Andreas Dilger Principal Software Engineer Cluster File Systems, Inc.
Hi, Here is the output of syslog Oct 4 09:36:20 mds02 kernel: loop: loaded (max 8 devices) Oct 4 09:36:20 mds02 kernel: Lustre: Acceptor stopping Oct 4 09:36:21 mds02 kernel: Lustre: Removed LNI 192.168.3.253@tcp Oct 4 09:36:38 mds02 kernel: Lustre: 7954:0:(module.c:382:init_libcfs_module()) maximum lustre stack 8192 Oct 4 09:36:38 mds02 kernel: Lustre: OBD class driver Build Version: 1.4.7-19691231170000-PRISTINE-.testsuite.tmp.boulder.lbuild-boulder.BUILD.lustre-kernel-2.6.9.lustre.linux-2.6.9-42.EL_lustre.1.4.7smp, info@clusterfs.com Oct 4 09:36:38 mds02 kernel: Lustre: Added LNI 192.168.3.253@tcp [8/256] Oct 4 09:36:38 mds02 kernel: Lustre: Accept secure, port 988 Oct 4 09:36:38 mds02 kernel: kjournald starting. Commit interval 5 seconds Oct 4 09:36:38 mds02 kernel: LDISKFS FS on hda3, internal journal Oct 4 09:36:38 mds02 kernel: LDISKFS-fs: mounted filesystem with ordered data mode. Oct 4 09:36:38 mds02 kernel: LustreError: 8157:0:(handler.c:2059:mds_postsetup()) No profile found: mds01 Oct 4 09:36:38 mds02 kernel: LustreError: 8257:0:(obd_config.c:288:class_cleanup()) Device 1 not setup What can I make from that? On Tue, 2006-10-03 at 23:23 -0600, Andreas Dilger wrote:> On Oct 04, 2006 00:34 +0200, Itamar Ofek wrote: > > I''m using Centos 4.4 (EL 4 clone ) with stock rpm from lustre. > > trying just to initially setup an mds > > I have got the ost''s up,however the mds fail to initialize > > > > suggestions ? > > > > MDS failed to start. Check the syslog for details. > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > > Cheers, Andreas > -- > Andreas Dilger > Principal Software Engineer > Cluster File Systems, Inc. >