Details
-
Bug
-
Resolution: Fixed
-
Minor
-
Lustre 2.6.0
-
None
Description
setupall()->start ost called at the end of large-lun.sh
fails to start the ost devices which leads test_1() and
test_2() fail if run separately.
The reasons are:
1) osts data are destroyed by test_1()->llverdev
2) osts are reformatted by test_2, but mds config not changed
To reproduce:
# sh llmount.sh # ONLY=1 sh large-lun.sh
xtrace:
+ true + setupall ... + start ost1 /dev/vdb -o user_xattr ... + /usr/bin/pdsh -R ssh -S -w fre806 '(PATH=$PATH:/usr/lib64/lustre/utils:/usr/lib64/lustre/tests:/sbin:/usr/sbin; cd /usr/lib64/lustre/tests; LUSTRE="/usr/lib64/lustre" sh -c "mkdir -p /mnt/ost1; mount -t lustre -o user_xattr /dev/vdb /mnt/ost1")' + sed 's/^fre806: //' fre806: mount.lustre: /dev/vdb has not been formatted with mkfs.lustre or the backend filesystem type is not supported by this tool pdsh@fre807: fre806: ssh exited with exit code 19 + return 19
# NOFORMAT=: sh llmount.sh
Checking servers environments
Checking clients fre807 environments
Loading modules from /usr/lib64/lustre/tests/..
detected 2 online CPUs by sysfs
...
Starting ost1: -o user_xattr /dev/vdb /mnt/ost1
fre806: mount.lustre: /dev/vdb has not been formatted with mkfs.lustre or the backend filesystem type is not supported by this tool
pdsh@fre807: fre806: ssh exited with exit code 19
[root@fre807 tests]#