Amanda-Users

Re: tapetype question

2006-08-01 19:20:37
Subject: Re: tapetype question
From: Jon LaBadie <jon AT jgcomp DOT com>
To: amanda-users AT amanda DOT org
Date: Tue, 1 Aug 2006 18:11:50 -0400
On Tue, Aug 01, 2006 at 01:53:45PM -0400, McGraw, Robert P. wrote:
> Presently I am using an SDLT220 tape drive with hardward compression turn
> on. I have a new LTO2 tape drive and I really do not want to run the
> amtapetype program if someone else has already done this. 
> 
> Here is my LTO2 drive:
> 
>       Product Type: Tape Drive
>       Vendor ID: 'HP      '
>       Product ID: 'Ultrium 2-SCSI  '
>       Revision: 'F63Z'
>       Attached Changer: No
>       SerialNumber: 'HUL5M02595'
>       MinBlock:1
>       MaxBlock:16777215
>       Ready: yes
> 
...
> 
> I am running Solaris 10 on the amanda server. I ran the first part of
> tapetype with /dev/rmt/1n(c) where "n" is no rewind and "c" is compression.
> I did not specify the "c" or "u" for compression, yet I got "WARNING: Tape
> drive has hardware compression enabled".
> 
...
> 
> 
> Any ideas on how to turn off compression?     
> 

I'm assuming things haven't changed drastically from Sol9.

Do you also have some other "1n" devices in /dev/rmt?
For example, do you have 1ln (low), 1mn (medium), 1hn (high)?
One of these `may' be the drive with compression turned off.

BTW devices "c" (compressed) and "u" (ultra) are always the
same.  And the names for the letters don't really mean anything
except convention.  "l" could be high density compressed
while "u" could be low density uncompressed.  Not often set up
that way, but who knows.  They could also all be the same.  The
device without letters 1n or just 1, matches one of the lettered
devices as the "default".  But which one is controlable.

These devices are controlled by driver settings specified
in /kernel/drv/st.conf.  During boot an id string is read from
the device.  It can be found in the massive output of prtconf,
the '-D' option I think.  That id string gets matched with one
in st.conf.  My Sol9 st.conf doesn't have an LTO entry so I
can't demo it exactly.  Sometimes you have to read the drive
manufacturers docs to find their recommendation for the lines.
But here is a line used by my old HP DDS3 6x24 changer.

"HP      C1557A",   "HP DDS-3 4mm DAT loader",   "HPdds3";

The first string is what is matched, spaces are important.
The last string (HPdds3) is used to get another config entry
later in the file.  Different manufacturers may output different
strings but use the same config entry.  My config entry is:

HPdds3 = 1, 0x34,    0, 0x0D639, 4, 0x00, 0x13, 0x24, 0x03, 0;

I once figured out what each of the first couple of fields were,
with various docs you can break 0xD639 to the bits turned on and
off and their meaning to the device driver.  But the last 6 fields
are what we care about.

  4     how many devices with letters will be created (4 max)
  0x00  some code to the device and driver for the "l" device
  0x13           ditto                             "m" device
  0x24           ditto                             "h" device
  0x03           ditto                      c" and "u" devices
  0     which of the 4 devices matchs the default (no letter) device.
        In this case it is the "l" which is the default.  I modified
        it from HP's recommended "3" (the c/u device) because I wanted
        it to be no compression.


On the net I found a suggested entry for LTO2 - I have no idea
if it is correct or reasonable.

ULTRIUM   = 1, 0x36,  0,  0xd639, 4, 0x00, 0x00, 0x00, 0x40, 3;

This one creates all the 4 of the lettered devices, but note that
l, m, and h are all the same (0x00).  With this entry, the default,
unlettered device is 3, matching the "c" and "u" devices.

I'm guessing the 0x00 devices are compression off, but don't
know that for certain.

At least you have a starting point to investigate.

-- 
Jon H. LaBadie                  jon AT jgcomp DOT com
 JG Computing
 4455 Province Line Road        (609) 252-0159
 Princeton, NJ  08540-4322      (609) 683-7220 (fax)

<Prev in Thread] Current Thread [Next in Thread>