boot failure - coreadm.conf read only file ???

[760613]

Hi!!!

I've got a Sun Netra X1 server running Solaris 9,
When rebooting I got this strange message:

FAILURE:
failed to open /etc/coreadm.conf: Read-only file system
INIT: Cannot create /var/adm/utmpx
INIT: failed write of utmpx entry:" "
INIT: failed write of utmpx entry:" "
INIT: SINGLE USER MODE

Type control-d to proceed with normal startup,
Type control-d to proceed with normal startup,

I login in Single-User Mode, and did this steps:
# mount -o rw,remount /

Control-d - to bring up the system multi-user.

#fsck -y /
...
FILE SYSTEM STATE IN SUPERBLOCK IS WRONG; FIX? yes
...

Reboot the system, but still see my error message.

What do I need to get off this failure,

I ve not modified the vfstab file, and its clean I've checked it more then twice! :'(

please helpppppp me!

 

[comtec17]

What was the last thing you did before this error occurred?

 

[760613]

Hi comtec,

>What was the last thing you did before this error occurred?

ok reset
I boot the machine everything went fine.
Later the only thing I ve done was sharing a directory
I added a share line in dfstab.
nfs.server start

And before leaving my work I shutdown (init 5) the server, and day after I started my server (lom poweron) and I recieved this message.

I dont know what the heck it could be.

 

[760613]

Whats the problem, am I the only one in the entire solaris world getting this error ??? plz someone, heeeeeeeeeeeeeeelp!

 

[thomgreen]

Hey 760613,
I have been having the same problem. have you found a solution. My error looks more like this:

/
INIT: Cannot create /var/adm/utmpx
INIT: failed write of utmpx entry:" "
INIT: failed write of utmpx entry:" "
INIT: SINGLE USER MODE

Type control-d to proceed with normal startup

 

[Annihilannic]

Try booting off CD-ROM and performing the fsck. You should never use fsck on a mounted filesystem.

Annihilannic.

 

[thomgreen]

Thanks for the response. I tried that and rebooted and still received the same error on boot-up. any ideas?

 

[Annihilannic]

Is your /var on a separate filesystem or on /?

If it's separate, have you checked both?

Are your disks mirrored? Are they working? Try testing them in format (use analyze / read).

Annihilannic.

 

[thomgreen]

/var is not a separate file system but off /
Disks are not mirrored. Just one disk in an Ultra 10. I will run format tonight.

I boot up from disk and am able to mount and access all mount points.

 

[Annihilannic]

Also if you could capture the boot messages from earlier on in the boot process that might be useful... there should be something that indicates why it's failing to remount root.

Annihilannic.

 

[bi]

thomgreen: If you are able to mount and access all mount points, do an ls -ld on /var and make sure the ownership and permissions are correct (root:sys, 775). Then do the same thing for /var/adm (root:sys, 775).

If ownerships and permissions of these two directories are OK, check to see if you have enough space on the mount point where /var is (either / or /var).

760613: you should also check ls -ld /. but your problem seems more complicated.

 

[thomgreen]

I checked the permissions on the /var/adm/utmpx file. the permissions were correct. I actually deleted and recreated the file and rebooted - didn't work. I then copied the utmpx file from disk to /var/adm as well... Still recieved the same error.

I will check the permissions on the directories tonight when I am able to access the machine. I have searched the error on google and sunsolve. Have seen some almost exactly the same except the first line of the error. Many have seen failure notices such as:
failed to open /etc/coreadm.conf: Read-only file system
but mine just has a:
/

 

[ddlldd2000]

Hey thomgreen,

Your problem is that you can't mount /var so you can't use /var/adm. Use "Ctrl+d" to single user mode, then you can fsck -y /dev/c0t0d0s3 to check and correct /var file system. After this, use "exit" to continue nornal boot. BTW, if you use diffirent slice for /var, just change the c0t0d0s3 to the real one.

 

[eileeeen]

I am having the same problem! The guy that used to maintain this computer left a while ago, but this is the first time I've come across a problem that I really can not figure out. Here's what happened: I rebooted, and it hung with a message saying:
"Setting default IPv4 interface for multicast: add net 224.0/4: gateway akers bind: Address already in
use."
Most likely an NFS problem, right? So I emailed the guy that used to maintain this computer (akers) and he says:
"try logging in as single-user and commenting out all references in vfstab".
In retrospect, I think he meant comment out SOME things in vfstab. I commented out just about everything. Now when I reboot, even from disk, I get these kinds of error messages: "/sbin/rcS: /etc/dfs/sharetab: cannot create
failed to open /etc/coreadm.conf: Read-only file system", etc.
I can't vi into any files now, and "755 chmod" doesn't seem to work. Everything's read-only, and I can't change the permissions, as far as I can tell.

HELP would be much appreciated! Thanks!

 

[z28sarefun]

Eileeen, when I see a "Address already in use" Error the first thing that comes to mind is an IP address conflict. With the machine down (ok prompt or probably single user mode) try pinging it's IP address. If the ping comes back someone else is using your IP addy.

B