ADSM-L

How to set up 3494 w/3590s

1996-03-19 14:35:03
Subject: How to set up 3494 w/3590s
From: Richard Sims <rbs AT ACS.BU DOT EDU>
Date: Tue, 19 Mar 1996 14:35:03 -0500
IBM documentation on setting up a 3494 containing 3590 drives is awfully
sketchy.  The ADSM documentation says to follow all instructions in the
Installation and User's Guide which comes with the 3494 (manual GC35-0154-1).
I followed Chapter 23 to set up the Library Device Driver, no problem.  What
is lacking is definitive guidance on dealing with the 3590 tape drives.  Page
200 subsection "Configuring the IBM 3490E Tape Drive" omits any mention of
3590 tape drive configuration, which seems like an oversight in version
upgrading of the manual from former support of only 3490 drives.  (If nothing
need be done is should definitively say so.)  For completeness at least I
performed an installation of the Atape driver, which seemed to recognize the
drives I had and established the proper rmt*, ascsi, and vscsi devices.  If I
attempt to follow the instructions thereafter on p.13 and, after running
cfgmgr and rebooting, I do 'smit tapes', there is no sign of 3590 tape drives
in the selection menus.  I get the impression that this is the way it is
supposed to turn out (though it be counterintuitive) because I can employ the
'tapeutil' command and nicely get reasonable info from the drives, including
them saying that their max blksize is 2MB, meaning that there is no need to
perform and instAtape procedure on p.15.

So I'm seeking clarification on this, if anyone can supply info.

As an aside: My CE and I discovered the hard way that the lmcpd daemon, as
supplied on diskette (ver. 2.4.1.00) is defective.  Attempting to utilize the
mtlib command after reboot results in an error message about device not
found. And if the 3494 reboots attempting to access it yields a message saying
that the device is offline, though the 3494 is in an online state.  (This
resulted in much wasted debugging time.)  Killing and restarting the daemon is
a circumvention.  (There seems to be no documentation on any signals that it
might be sent to jog it.)  IBM knew that it was defective, but did not
communicate with CEs or registered customers working with this equipment.  IBM
could do better.  My CE is obtaining a fixed version of lmcpd.

      Richard Sims, Boston University OIT
<Prev in Thread] Current Thread [Next in Thread>
  • How to set up 3494 w/3590s, Richard Sims <=