Hi,
I've tried to move VM's from a datastore to another.
It worked fine.
But some VM left some files and directories behind.
For example, I've found a vmxf file not modified since 2 years, and it can't be deleted.
With command vkmfstools -D I have this result :
vmkfstools -D /vmfs/volumes
/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS2.vmxf
Lock [type 10c00001 offset 57563136 v 945, hb offset 3723264
gen 627, mode 0, owner 00000000-00000000-0000-000000000000 mtime 873810
num 0 gblnum 0 gblgen 0 gblbrk 0]
Addr <4, 99, 19>, gen 53, links 1, type reg, flags 0, uid 0, gid 0, mode 600
len 3232, nb 1 tbz 0, cow 0, newSinceEpoch 1, zla 2, bs 8192
It's in mode 0 so not locked, and I also can't find any MAC.
the rm command :
rm NAGIOS2.vmxf
rm: can't remove 'NAGIOS2.vmxf': No such file or directory
I've got another symptom an a directory that apperas empty and that I can't remove :
rmdir logs2
rmdir: 'logs2': Directory not empty
ls logs2
ls: logs2/rhttpproxy.0.gz: No such file or directory
Has someone a solution ?
Regards.
Looks like the inode entry for the files could be missing. If your end goal is to remove (unmount) the datastore and don't really care about the files on it, unpresent it from the hosts on the storage array and reboot the hosts. Once the hosts are rebooted, they should not see this datastore anymore. On the vCenter unless there are any VM / VM Templates still registered from this datastore, perform a cluster level rescan once the hosts are rebooted and the datastore should not be visible on the vCenter anymore. This is a hard way to remove any datastore and should be performed as the last resort option.
Cheers,
Supreet
Hi,
You can find more information about this file with this command in SSH from ESXi.
# vmfsfilelockinfo -p /vmfs/volumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS2.vmxf
Thank you,
Can you run the command <ls -lah> from within the folder '/vmfs/volumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/' and 'logs2' and share the output?
Cheers,
Supreet
Hi,
thank you for your time !
Here's the result of the command :
ls -lah
total 1040
drwxr-xr-x 1 root root 420 Sep 30 18:38 .
drwxr-xr-t 1 root root 2.2K Oct 3 07:57 ..
-rw------- 1 root root 3.2K Jun 23 2017 NAGIOS2.vmxf
Marmotte, here's the result :
vmfsfilelockinfo -p /vmfs/v
olumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS2.vmxf
vmfsfilelockinfo Version 2.0
Looking for lock owners on "NAGIOS2.vmxf"
"NAGIOS2.vmxf" is not locked by any ESX host and is Free
Total time taken : 0.03462965099606663 seconds.
Thank you for answering me.
Have you tried to delete the *vmxf from commandline with
rm -f /vmfs/volumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS2.vmxf
if that does not work try
dd if=/dev/zero of=/vmfs/volumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS2.vmxf bs=1 count=1
rm -f /vmfs/volumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS2.vmxf
Hi continuum,
each command you wote me doesn't work.
rm -f /vmfs/volumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS2.vmxf
rm: can't remove '/vmfs/volumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS2.vmxf': No such file or directory
dd if=/dev/zero of=/vmfs/volumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS
2.vmxf bs=1 count=1
dd: can't open '/vmfs/volumes/559bd310-c01ae484-6591-a0369f6fcdc8/NAGIOS/NAGIOS2.vmxf': No such file or directory
Regards
I used the filenames and the path that you provided yourself - if that aint the correct full path - adjust it accordingly
That's the correct path and file.
It's weird to see the file with ls or in the vCenter Interface, and to see that when you try to delete it, it appears as an unexisting file !
Did you try from an other ESXi ?
I think something wrong with Datastore ID or this file doesn't exist really (Same as the spoon).
Thank you,
Marmotte, I tried from my second ESXi and same result.
And I want to remove this datastore, but I can' because vCenter say's it's in use !
Looks like the inode entry for the files could be missing. If your end goal is to remove (unmount) the datastore and don't really care about the files on it, unpresent it from the hosts on the storage array and reboot the hosts. Once the hosts are rebooted, they should not see this datastore anymore. On the vCenter unless there are any VM / VM Templates still registered from this datastore, perform a cluster level rescan once the hosts are rebooted and the datastore should not be visible on the vCenter anymore. This is a hard way to remove any datastore and should be performed as the last resort option.
Cheers,
Supreet
I think you can perform a rescan Storage Adapters and refresh Storage device as well.
Thank you.
For the moment, recan for the storage adapters and storage didn't work.
I think I'll try what SupreetK wrote !
I'll post the result later !
Hi,
Used the last solution.
I deleted the lun on my SAN, and restarted every host.
Everything seems to be good now.
Thank all for your help !