Printing problems after losing /opt

[Marshall McDougall]

I lost a disk last week with /home, /tmp,and /opt on it. I have recreated /home and /tmp on the root vol but there is insuffucient space for /opt there. Subsequently, I have restored /opt to a larger volume and created a link (ln -s /space3/opt /opt) to it. Unfortunately, my spool starts, the queues are shown, but I cannot make black marks on paper. I have deleted the queues and recreated them still to no avail. In my naiveity, I believed that make_recovery was backing up /opt, but /opt was on vg01. Bummer. This is an off support D350 running HP-UX 11.0, so replaceing the drive is going to take a while. How do I make sure I have all the elements to print? Any and all suggestions welcome. Thanks in advance.
Regards, Marshall

[Steven E. Protter]

There are soft links in /opt with many print spoolers that point to /var where print jobs are stored.

If those are not right the problem may be there.

SEP

[Marshall McDougall]

Thanks for the quick response Steven. Can you say which links may be necessary? Thanks.

Regards, Marshall

[A. Clay Stephenson · Accepted solution]

The only thing in the interface files are references to hpnp (for the older JetAdmin) or hpnpl (for the newer JetDirect). I assume that you have installed JetDirect which would be in /opt/hpnpl.

I would have a look in /var/spool/lp/log for clues. You may have some hpnpf's that need to be killed.

Lp is also very sensitive to correct ownership and group settings as well as setuid, setgid bit preservation. If you simply cp'ed the data (as opposed to a tar or cpio, you lost that metadata).

[Marshall McDougall]

A. Clay Stephenson hit the nail on the head. I initially did just rcp the whole /opt over. After your reply, I created a tarball and copied that instead. I had to recreate my queues again, but it worked just fine. Thank you very much.

Regards, Marshall McDougall