/dev/root full

[R.Suresh]

Dear all,

I am having D230 server,

In which the dmesg gives the error

vxfs: mesg 001: vx_nospace - /dev/root file system full (1 block extent)

vxfs: mesg 001: vx_nospace - /dev/root file system full (1 block extent)

My root file system size as of Now

Filesystem kbytes used avail %used Mounted on
/dev/vg00/lvol3 1536000 134285 1314369 9% /
/dev/vg00/lvol1 47829 25775 17271 60% /stand

Is there any problem?

Are /dev/vg00/lvol3 & /dev/root different?

If so How can I rectify this problem?

Thanks
R.Suresh

[Stefan Farrelly · Accepted solution]

Well, there looks like plenty of space on /. Check for full messages in /var/adm/syslog/syslog.log to see what date+time they actually happened, they may be very old messages.

[John Palmer]

This looks like an old message - you certainly don't have a problem now. If you are on HP-UX 11.00 check /var/adm/syslog/syslog.log for the same message - it will have a timestamp.

The kernel reports the root filesystem as '/dev/root' it is the same volume as your /dev/vg00/lvol3.

By the way, your root filesystem at 1.5Gb is the largest that I have ever seen using an LVM layout.

[CHRIS_ANORUO]

This is an old message. lvol3, root and "/" are the same thing. You can use this command to check for large files in your filesystems:
find / -xdev -size +1,000,000c -exec ll {} \;

[Shelby Dunlap]

I concur with the previous accessment that the message look old. /dev/root IS indeed /dev/vg00/lvol3 on your system. It usually will be called this when you've booted your system into LVM maintenance mode, and then rebooted back up normally. A bdf would then show you :

/dev/root 1536000 134285 1314369 9% /
/dev/vg00/lvol1 47829 25775 17271 60% /stand

To correct this, you simply remove /etc/mnttab and issue a "mount -a" to recreate the mnttab file.

As you can see from your bdf output, someone has already fixed this since it now shows

/dev/vg00/lvol3 1536000 134285 1314369 9% /
/dev/vg00/lvol1 47829 25775 17271 60% /stand

furthermore solidifying the idea that these are indeed old error messages in the message buffer.

[John Palmer]

The kernel always appears to log messages pertaining to the root file system in syslog as '/dev/root'.

The /dev/root reported by bdf after maintenance mode is another issue.

[Bettina Ofner]

Check /dev/rmt for core files, or for tar files. Maybe somebody made a mistake using tar.
Bettina

[MANOJ SRIVASTAVA]

Sometimes df on and < partition name eg /dev/vg00/lvol3 > will give different results as the thread process to sync the background command doesnt do it on input source of both . There is a Patch to rectify this .

Also th message can be a old one and init q would give the right result .