Dear All,
I am working on AIX 5.2, and I have allocated 3072M disk space to /home filesystem. Now system has no free space at /home filesystem. But when I checked data on /home, I have only 900M data. Please if anybody can tell me about this difference and how this can be resolved.
Thank You.
Aqeel Anwar
Please run a df command and paste the result here. Also take a listing of rootvg (where I trust /home resides) use the command lsvg -l rootvg. Also lsvg rootvg. Paste all out put on your reply.
# lsvg rootvg
VOLUME GROUP: rootvg VG IDENTIFIER: 000cec8e00004c00000001042ce983d7
VG STATE: active PP SIZE: 64 megabyte(s)
VG PERMISSION: read/write TOTAL PPs: 542 (34688 megabytes)
MAX LVs: 256 FREE PPs: 5 (320 megabytes)
LVs: 17 USED PPs: 537 (34368 megabytes)
OPEN LVs: 13 QUORUM: 2
TOTAL PVs: 1 VG DESCRIPTORS: 2
STALE PVs: 0 STALE PPs: 0
ACTIVE PVs: 1 AUTO ON: yes
MAX PPs per PV: 1016 MAX PVs: 32
LTG size: 128 kilobyte(s) AUTO SYNC: no
HOT SPARE: no BB POLICY: relocatable
# lsvg rootvg -l
rootvg:
LV NAME TYPE LPs PPs PVs LV STATE MOUNT POINT
hd5 boot 1 1 1 closed/syncd N/A
hd6 paging 64 64 1 open/syncd N/A
hd8 jfslog 1 1 1 open/syncd N/A
hd4 jfs 2 2 1 open/syncd /
hd2 jfs 24 24 1 open/syncd /usr
hd9var jfs 2 2 1 open/syncd /var
hd3 jfs 9 9 1 open/syncd /tmp
hd1 jfs 200 200 1 open/syncd /home
hd10opt jfs 1 1 1 open/syncd /opt
lv00 jfs 47 47 1 closed/syncd /ptf
loglv00 jfs2log 1 1 1 open/syncd N/A
fslv00 jfs2 48 48 1 open/syncd /bea
lvora jfs2 81 81 1 open/syncd /oracle
ptcl2_redo01 raw 8 8 1 closed/syncd N/A
ptcl2_redo02 raw 8 8 1 open/syncd N/A
ptcl2_redo03 raw 8 8 1 closed/syncd N/A
lvbill jfs2 32 32 1 open/syncd /bill
# df -m
Filesystem MB blocks Free %Used Iused %Iused Mounted on
/dev/hd4 128.00 92.93 28% 2018 4% /
/dev/hd2 1536.00 24.34 99% 31878 9% /usr
/dev/hd9var 128.00 12.62 91% 610 2% /var
/dev/hd3 576.00 481.65 17% 279 1% /tmp
/dev/hd1 3072.00 20.90 100% 15126 2% /home
/proc - - - - - /proc
/dev/hd10opt 64.00 28.31 56% 626 4% /opt
/dev/fslv00 3072.00 1208.51 61% 19071 7% /bea
/dev/lvora 5184.00 412.23 93% 16058 15% /oracle
/dev/lvoradata 37120.00 1488.08 96% 26 1% /oradata
/dev/lvbill 2048.00 1099.20 47% 3637 2% /bill
/dev/lvjfdata 29312.00 8252.37 72% 27601 2% /jfdata
# du -sm /home
878.23 /home
This does not add up!
Your rootvg shows that the PP size is 64M and /home mounted on rootvg has 200 PPs. Therefore 200 * 64 = 12800 means you should have a filesystem which has 12.8GByte. Yet your df -m shows you have a /home filesystem which has only 3GByte. Something very peculiar here. The only thing I can see which is questionable is the Quorum being set to 2 when this is a single disk VG (I trust). Please run the command lspv | grep rootvg and paste the result here. What is this machine used for Oracle or something similar? I noticed the raw devices. Are there any other disks?
# lspv|grep rootvg
hdisk0 000cec8e2ce9663e rootvg active
yes we have total 3 HDs installed in this system.
# lspv
hdisk0 000cec8e2ce9663e rootvg active
hdisk1 000cec8ec654d5ba datavg active
hdisk2 005b68ec33ac7a01 datavg active
We are using this server as database (oracle) and application server.
This still makes no sense. I have 2 questions and 1 request. The question when was the machine re-booted last? Are there any errors in the errpt? The request; is it possible to have an ls -l listing of /home? You are 9G adrift somewhere!
Are you using any type of fs manager that pools your data and then splits it into your mount points?
ie.. on our old digital platform we used advfs which pooled say 90gb but we could allocate 30gb across 3 mount points.
We found a reporting bug.. in that if one of the file systems reached 100%, even after cleaning up data, the df command did not report the free space.
Its possible that the native OS could have a similar reporting bug.
Last time system rebooted 57 days back.
yes, there are errors in errpt but no error is related to /home filesystem.
# pwd
/home
# ls -l
total 27
drwxr-xr-x 2 bin system 512 Mar 10 2006 TT_DB
drwx------ 8 appsvc dba 512 Aug 18 16:27 appsvc
drwxr-xr-x 10 bill1 staff 1024 Nov 15 16:46 bill1
drwxr-xr-x 2 brio staff 512 Jan 15 2006 brio
drwxr-xr-x 35 cash staff 2048 Nov 15 15:51 cash
-rw-r--r-- 1 root system 2696 Nov 14 09:34 core
drwxr-xr-x 3 guest usr 512 Jun 14 15:54 guest
drwxr-xr-x 2 root system 512 Aug 18 11:13 long
drwx------ 2 root system 512 May 30 2005 lost+found
drwxr-xr-x 11 media staff 1024 Nov 06 09:33 media
drwxr-xr-x 5 oracle system 512 May 31 10:48 oracle
drwxr-xr-x 2 saad staff 512 Aug 18 09:19 saad
drwxr-xr-x 2 sco staff 512 Jan 17 2006 sco
drwxr-xr-x 10 sett staff 512 Jul 04 13:14 sett
drwxr-xr-x 2 bin staff 512 Mar 21 2006 veritas
drwxr-xr-x 2 root system 512 May 26 17:05 zxccdata
drwx------ 19 zxin10 dba 512 Jun 29 18:41 zxin10
I don't use AIX so take this with a grain of salt... If I understand, hd1 is a logical volume that contains /home but it appears that /home is too small. With other OS's, taking HP-UX as an example, there is an lvextend command to grow a logical volume and there is an extendfs command to grow the filesystem. And it is possible to run the former and forget the latter, thus creating this very situation. Maybe that happened here?
Thanks to all for help, this problem has been fixed now.
Acutallly rather than using chfs -a size+XXX /home to increase filesystem size, the command extendlv has been used instead. i.e. extendlv XX hd1
The output of lsfs states that the filesystem size is blocks is: 6291456 blocks
A filesystem of 1PP of 64MB has 131072 512 byte blocks, so the following command showed that the filesyetem is only using 48 of the 200 PP.
#dc
6291456 131072 / p
48
so, the problem is fixed by using following commands to use all the space that had been allocated:
chlv -x 200 hd1
chfs -a size=12800M /home
Thanks