I was trying to delete and recreate a filesystem, thinking it would be a quick way of deleting all of the files off of it, and I fear that I messed up. Here is the original entry from /etc/filesystems:
data01:
dev = /dev/fslv00
vfs = jfs2
log = /dev/loglv00
mount = true
options = rw
account = false
And the output from df:
/dev/fslv00 133169152 0 100% 4639433 60% /data01
I first ran rmfs /data01 to delete the filesystem, then I tried the following:
[root@ibm: /]# mkfs -V jfs2 -o name=/data01 /dev/fslv00
mkfs: destroy /dev/fslv00 (yes)? y
mkfs: 0506-214: /dev/fslv00: A file or directory in the path name does not exist.
This didn't work so I tried crfs instead, but I don't think I was using that properly either:
[root@ibm: dev]# crfs -m /data01 -d /dev/fslv00 -v jfs2
0516-306 getlvodm: Unable to find fslv00 in the Device
Configuration Database.
/usr/sbin/getlvodm: A flag requires a parameter: y
Usage: /usr/sbin/getlvodm [-a LVdescript] [-B LVdescript] [-b LVid] [-c LVid]
[-C] [-d VGdescript] [-e LVid] [-F] [-g PVid] [-h] [-j PVdescript]
[-k] [-L VGdescript] [-l LVdescript] [-m LVid] [-p PVdescript]
[-r LVid] [-s VGdescript] [-t VGid] [-u VGdescript] [-v VGdescript]
[-w VGid] [-y LVid] [-G LVdescript]
crfs: 0506-923 Cannot get logical volume type from odm.
I am just trying to recreate my old /data01 filesystem with "normal" settings. The man pages for mkfs/crfs are a little confusing.. Any help would be appreciated.
Thank you Scott! I created a new logical volume from my volume group data01, like you said:
[root@ibm: dev]# mklv data01 65G
lv00
[root@ibm: dev]# lslv lv00
LOGICAL VOLUME: lv00 VOLUME GROUP: data01
LV IDENTIFIER: 00c3e63a00004c000000013535af9d5f.3 PERMISSION: read/write
VG STATE: active/complete LV STATE: closed/syncd
TYPE: jfs WRITE VERIFY: off
MAX LPs: 520 PP SIZE: 128 megabyte(s)
COPIES: 1 SCHED POLICY: parallel
LPs: 520 PPs: 520
STALE PPs: 0 BB POLICY: relocatable
INTER-POLICY: minimum RELOCATABLE: yes
INTRA-POLICY: middle UPPER BOUND: 32
MOUNT POINT: N/A LABEL: None
MIRROR WRITE CONSISTENCY: on/ACTIVE
EACH LP COPY ON A SEPARATE PV ?: yes
Serialize IO ?: NO
Then I tried to create the new filesystem:
[root@ibm: dev]# mkfs -V jfs2 -o name=/data01 /dev/lv00
mkfs: destroy /dev/lv00 (yes)? y
File system created successfully.
68155156 kilobytes total disk space.
Device /dev/lv00:
Standard empty filesystem
Mount point: /data01
Size: 136310312 512-byte (DEVBLKSIZE) blocks
However when I ran df my new filesystem did not show up:
Also my /etc/filesystems file does not have an entry for /data01.
Did I do something wrong? Was I supposed to set an auto-mount option or something?
When I look at the volume group information, my new logical volume shows up but it says N/A for mount point, was I supposed to do something when creating the new LV?
[root@ibm: /]# lsvg -l data01
data01:
LV NAME TYPE LPs PPs PVs LV STATE MOUNT POINT
lvdata01 jfs 1 1 1 closed/syncd N/A
loglv00 jfs2log 1 1 1 closed/syncd N/A
lv00 jfs 520 520 1 closed/syncd N/A
My AIX is very rusty! Perhaps you should have used crfs, actually, with the -A yes option to automount the filesystem. Still, no worries. You can just copy-paste the previous entry into /etc/filesystem, changing the LV name and any other details as appropriate, and then mount the filesystem manually.
/data01:
dev = /dev/lv00
vfs = jfs2
log = /dev/loglv00
mount = true
options = rw
account = false
(I added the /, as I assume it's missing)
It's not clear where lvdata01 came from. Did you just create that too?
If the filesystem /data01 is for database data, you might consider adding noatime to the options.
Hm I'm not sure where lvdata01 came from either, and nope we won't be using this for database data
Thank you so much for the help! I'm not super familiar with how filesystems are set up on AIX, and I wasn't sure if I was even allowed to modify /etc/filesystems but now everything is mounted properly. Cheers
Sure, you can edit the file if you're careful. Normally commands like chfs, smit(ty), etc. would update it for you, but as the whole data01 stanza was missing, it's easier just to add it manually.