Shell Programming and Scripting

Also, the '*' output by the NAS probably means that the share is not tied to a particular user, i.e., share can be accessed by all-comers.

Cool..

So now?:
8. Change directory to the top of the hard disk (empty) root

Code:

# cd \

Code:

# ls -l

should see-

Code:

Lost+Found

9. 'ufsrestore' the backup from the NFS storage to the root hard disk

Code:

# ufsrestore rf /mnt/<dumpfile>

Yes, exactly. The only thing to check is the exact path of your dumpfile....

Code:

# ufsrestore rf /mnt/<dumpfile>

assumes your dumpfile is in the, now mounted, /mnt directory. Is it? Or are there other directories below the share?

e.g.

Code:

# ufsrestore rf /mnt/dir1/dir2/<dumpfile>

Checkout the exact path to your dumpfile, then change directory to your empty hard disk root filesystem, e.g.

Code:

# cd /a

and then restore your root filesystem from the dump.

Repeat, just check that you know the dumpfile path.

And......when successfully restored, be careful when rebooting because this clone will come up with the same ip address as the production box unless you do something about it. Could cause network chaos.

OK Thanks again!

so I did:

Code:

cd /a
 ufsrestore rf /mnt/dir1/dir2/<dumpfile.ufs>

and got many error messages such as:

Code:

Warning: /.home/<dir>/<dir>...... No such file or directory

So I interrupted it. Is this normal to see? Nothing was in

Code:

/a

No, that ain't right!!!!!

The only error I would expect to see is a warning that Lost+Found already exists.

Ok, let's recap exactly what we're doing here. Unlike at the start of this thread, you know all of this stuff now so you can very quickly check.

1. This is a Disaster Recovery (DR) routine so we start with an empty machine. Boot from DVD into single user to get a # prompt.
2. Use 'format' to label, partition (VTOC) our slices, and 'newfs' at least the first slice to take our restored root filesystem.
3. We have two mountpoints available on the DVD; /a and /mnt. We are going to restore from a dumpfile mounted on /mnt to our new hard disk (empty) root filesystem mounted on /a. So first mount slice one of our new empty filesystem on /a. Looking in here now, all we see is Lost+Found.
4. Mount our remote NAS dumpfile on /mnt. Looking in /mnt now we see our NAS files.
5. Now change directory back to /a; check again that we only have Lost+Found in here. Issue 'ufsrestore' to tip the whole contents of our dumpfile into this directory. Only error should be a warning that Lost+Found already exists.

That's where we are up to right now so what are you doing wrong????

EDIT: I omitted to say that, at some point, you need to plumb, configure and up a network interface in order to reach the NAS over the LAN.

Without redoing steps 1 through 4 yet, I can say that in step 5, I do not see anything in

Code:

/a

, not even

Code:

Lost+Found

. I believe it went away when I was trying to find

Code:

mnt

and did

Code:

cd /

to see

Code:

mnt

.