I use lustre 1-0-4 on i386 platform. I have a problem with mds recovery. I have configured 5 node with following configuration file. My LDAP server is on station8. /home is shared NFS filesystem between all nodes. lmc -o config.xml --add net --node station5 --nid station5 --nettype tcp lmc -m config.xml --add net --node station8 --nid station8 --nettype tcp lmc -m config.xml --add net --node station9 --nid station9 --nettype tcp lmc -m config.xml --add net --node station10 --nid station10 --nettype tcp lmc -m config.xml --add net --node station11 --nid station11 --nettype tcp lmc -m config.xml --add mds --node station5 --mds mds1 --fstype ext3 --dev /home/mds1 --size 50000 --group LMDS lmc -m config.xml --add mds --node station8 --mds mds1 --fstype ext3 --dev /home/mds1 --size 50000 --group LMDS lmc -m config.xml --add lov --lov lov1 --mds mds1 --stripe_sz 65536 --stripe_cnt 0 --stripe_pattern 0 lmc -m config.xml --add ost --node station9 --lov lov1 --ost ost1 --fstype ext3 --dev /tmp/ost1 --size 100000 lmc -m config.xml --add mtpt --node station10 --path /mnt/lustre --mds mds1 --lov lov1 lmc -m config.xml --add mtpt --node station11 --path /mnt/lustre --mds mds1 --lov lov1 Then, I configured every node. [station9] lconf –ldapurl –v --reformat ldap://station8 --node station9 --config config [station5] lconf --ldapurl ldap://station8 --node station5 --config config -v --reformat [station10] lconf --ldapurl ldap://station8 --node station10 --config config --timeout=100 --lustre_upcall=/home/mdsrecover -v –reformat [station11] lconf --ldapurl ldap://station8 --node station11 --config config --timeout=100 --lustre_upcall=/home/mdsrecover -v –reformat Then I wrote a script at /home address, named mdsrecover which contained these instructions: /home/lustre-1.0.4/utils/lactive --ldapurl ldap://station8 --config config --group LMDS --active station8 lconf -v --ldapurl ldap://station8 --config config --node station8 Next, I cleaned the first MDS server using following instruction on station5: lconf --ldapurl ldap://station8 --config config --force --cleanup --node station5 Then, I executed mdsrecover to recover MDS server on station8, but this message appeared: mds1_UUID not active Do you know what is wrong? -- mojtaba soltani