I have a 3 system test configuration. 1 machine with with a mds, 1
machine with 2 ost''s and 1 machine that will be the client. My first
test was to run a single system test. It ran great and was very easy to
setup. My next test included separating the three components. I
generated a xml file named alpha.xml using lwizard. I copied the
alpha.xml file to each of the machines and used "lconf -v --reformat
--gdb alpha.xml" on the OST machine first, then ran the same command on
the mds machine. Both the OST and mds machine had no errors and seemed
to be ready to have a client mount a lustre file system. To verify that
each of the two machines were running the lustre components correctly I
observed the output of "find /proc/fs/lustre" on both machines and it
looked fine. The last step was to go to the client machine and run
"lconf --node client alpha.xml". Everything seemed to run fine until
the
last stage and I received
mount failed: /mnt/lustre : mount: wrong fs type, bad option, bad
superblock on alpha,
or too many mounted file systems
Note: The machine that originally was used for the system test
configuration was the client machine. and the file system looks like:
[root@lustre alpha]# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/hda2 8.6G 1.8G 6.4G 22% /
/dev/hdb1 4.0G 130M 3.7G 4% /b
/dev/hda1 101M 15M 81M 16% /boot
/dev/hdd1 3.0G 53M 2.8G 2% /d
none 141M 0 141M 0% /dev/shm
Does anyone have an idea? Thanks in advance.
Gary McElwee
This is the complete trace:
[root@lustre alpha]# lconf -v --node client alpha.xml
configuring for host: [''client'']
add_local NET_client_tcp_UUID
find_local_routes: []
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_client_tcp NET_client_tcp_UUID
Service: ldlm ldlm ldlm_UUID
Service: mountpoint MNT_client MNT_client_UUID
+ sysctl portals/debug_path /tmp/lustre-log-lustre
+ /usr/sbin/lctl modules > /tmp/ogdb-lustre
Service: network NET_client_tcp NET_client_tcp_UUID
NETWORK: NET_client_tcp NET_client_tcp_UUID tcp client 988
+ /usr/sbin/lctl
network tcp
mynid client
quit
Service: ldlm ldlm ldlm_UUID
Service: mountpoint MNT_client MNT_client_UUID
OSC: OSC_lustre_OST_sc3_MNT_client 39875_lov1_4787d89a0d OST_sc3_UUID
+ /usr/sbin/lctl
add_uuid NID_sc3_UUID sc3 tcp
+ /usr/sbin/lctl
network tcp
send_mem 8388608
recv_mem 8388608
add_autoconn sc3 sc3 988 s
quit
OST_sc3_UUID active
+ /usr/sbin/lctl
attach osc OSC_lustre_OST_sc3_MNT_client 39875_lov1_4787d89a0d
quit
+ /usr/sbin/lctl
cfg_device OSC_lustre_OST_sc3_MNT_client
setup OST_sc3_UUID NID_sc3_UUID
quit
OSC: OSC_lustre_OST_sc3_2_MNT_client 39875_lov1_4787d89a0d
OST_sc3_2_UUID
+ /usr/sbin/lctl
add_uuid NID_sc3_UUID sc3 tcp
+ /usr/sbin/lctl
network tcp
send_mem 8388608
recv_mem 8388608
add_autoconn sc3 sc3 988 s
quit
OST_sc3_2_UUID active
+ /usr/sbin/lctl
attach osc OSC_lustre_OST_sc3_2_MNT_client 39875_lov1_4787d89a0d
quit
+ /usr/sbin/lctl
cfg_device OSC_lustre_OST_sc3_2_MNT_client
setup OST_sc3_2_UUID NID_sc3_UUID
quit
LOV: lov1 39875_lov1_4787d89a0d mds1_UUID 0 1048576 0 0
[u''OST_sc3_UUID'', u''OST_sc3_2_UUID''] mds1
+ /usr/sbin/lctl
attach lov lov1 39875_lov1_4787d89a0d
lov_setup lov1_UUID 0 1048576 0 0 OST_sc3_UUID OST_sc3_2_UUID
quit
MDC: MDC_lustre_mds1_MNT_client 72292_MNT_client_40a04598a6 mds1_UUID
+ /usr/sbin/lctl
add_uuid NID_sc2_UUID sc2 tcp
+ /usr/sbin/lctl
network tcp
send_mem 8388608
recv_mem 8388608
add_autoconn sc2 sc2 988 s
quit
mds1_UUID active
+ /usr/sbin/lctl
attach mdc MDC_lustre_mds1_MNT_client 72292_MNT_client_40a04598a6
quit
+ /usr/sbin/lctl
cfg_device MDC_lustre_mds1_MNT_client
setup mds1_UUID NID_sc2_UUID
quit
MTPT: MNT_client MNT_client_UUID /mnt/lustre mds1_UUID lov1_UUID
+ mkdir /mnt/lustre
+ mount -t lustre_lite -o osc=lov1,mdc=MDC_lustre_mds1_MNT_client alpha
/mnt/lustre
MDC: MDC_lustre_mds1_MNT_client 72292_MNT_client_40a04598a6
+ /usr/sbin/lctl
ignore_errors
cfg_device $MDC_lustre_mds1_MNT_client
cleanup
detach
quit
+ /usr/sbin/lctl
ignore_errors
del_uuid NID_sc2_UUID
quit
+ /usr/sbin/lctl
ignore_errors
network tcp
del_autoconn sc2 sc2 s
quit
LOV: lov1 39875_lov1_4787d89a0d
+ /usr/sbin/lctl
ignore_errors
cfg_device $lov1
cleanup
detach
quit
OSC: OSC_lustre_OST_sc3_MNT_client 39875_lov1_4787d89a0d
+ /usr/sbin/lctl
ignore_errors
cfg_device $OSC_lustre_OST_sc3_MNT_client
cleanup
detach
quit
+ /usr/sbin/lctl
ignore_errors
del_uuid NID_sc3_UUID
quit
+ /usr/sbin/lctl
ignore_errors
network tcp
del_autoconn sc3 sc3 s
quit
OSC: OSC_lustre_OST_sc3_2_MNT_client 39875_lov1_4787d89a0d
+ /usr/sbin/lctl
ignore_errors
cfg_device $OSC_lustre_OST_sc3_2_MNT_client
cleanup
detach
quit
+ /usr/sbin/lctl
ignore_errors
del_uuid NID_sc3_UUID
quit
+ /usr/sbin/lctl
ignore_errors
network tcp
del_autoconn sc3 sc3 s
quit
mount failed: /mnt/lustre : mount: wrong fs type, bad option, bad
superblock on alpha,
or too many mounted file systems
--
Thank You,
Gary McElwee
www.scopular.com
gary@scopular.com
Ph (302) 584-7599
Fax (801) 761-6403
Gary McElwee wrote:> The last step was to go to the client machine and run > "lconf --node client alpha.xml". Everything seemed to run fine until the > last stage and I received > > mount failed: /mnt/lustre : mount: wrong fs type, bad option, bad > superblock on alpha, > or too many mounted file systemsWhen you are reporting any kind of Lustre problem, almost all of the interesting messages are on the console, in dmesg, or in /var/log/messages. In the future, please include any Lustre messages from each node. The most common problem of this type, however, is when your hostname is defined to be 127.0.0.1 instead of the real IP address. If you see: 127.0.0.1 localhost.localdomain localhost hostname in your /etc/hosts, this will cause a problem. Please make sure that the hostname points to the real IP address that you''re trying to use. If this does not resolve your issue, please send along the Lustre messages from the console, and we''ll work through it. Thanks-- -Phil