Is this an ignite bug?

 
George Liu_2
Frequent Advisor

Is this an ignite bug?

# /opt/ignite/bin/make_recovery -p -d /dev/rmt/0m
NOTE: /opt/ignite/bin/make_recovery(103): The make_recovery command will be replaced by
make_tape_recovery. Both commands are supported in this
release (Ignite-UX Revision 3.2). Please read the man page
make_tape_recovery(1M) for usage of the new command. In a
future release, make_recovery will be replaced by a script
that calls make_tape_recovery.


*** Previewing only ***
make_recovery(284): Writing to the tape failed. Please check the existence of the tape, or the tape is write protected, or the tape is bad.
make_recovery(210): Encountered a fatal error.
make_recovery(212): Please fix errors and re-try.
Cleanup
# ioscan -funC tape
Class I H/W Path Driver S/W State H/W Type Description
=====================================================================
tape 0 8/12/5.1.0 stape CLAIMED DEVICE HP C1533A
/dev/rmt/0m /dev/rmt/c1t1d0BEST
/dev/rmt/0mb /dev/rmt/c1t1d0BESTb
/dev/rmt/0mn /dev/rmt/c1t1d0BESTn
/dev/rmt/0mnb /dev/rmt/c1t1d0BESTnb
#

Tape drive can be accessed on SAM to do backup.

HPUX 10.20

Thanks,
10 REPLIES 10
Ted Ellis_2
Honored Contributor

Re: Is this an ignite bug?

since you are trying to use the default device, drop the -d from the command and let it get the defaults... you should go with the no-rewind device in any case... 0mn
Ted Ellis_2
Honored Contributor

Re: Is this an ignite bug?

that would be drop "-d /dev/rmt/0m" or use /dev/rmt/0mn

Ted
Deshpande Prashant
Honored Contributor

Re: Is this an ignite bug?

Hi
Use the "make_tape_recovery" command with the non-rewinding device.

Thanks.
Prashant.
Take it as it comes.
Hai Nguyen_1
Honored Contributor

Re: Is this an ignite bug?

George,

You should use the newer version, make_tape_recovery. The device file should have been /dev/rmt/0mn (you missed an n at the end.). You do not need to mention the device file unless you want to specify one like /dev/rmt/1mn.

Hai
George Liu_2
Frequent Advisor

Re: Is this an ignite bug?

Problem is due to the /dev/rmt/0m or /dev/rmt/0mn that were manually updated recently and need to reboot the host to make the system to recognize them.

Anyway, the default tape drive is /dev/rmt/0m.
make_tape_recovery was called by make_recovery.

Thanks.
harry d brown jr
Honored Contributor

Re: Is this an ignite bug?


George,

Install the newest version for 10.20: A.3.7.95

http://software.hp.com/products/IUX/download.html

Plus, use the custom patch manager to make sure you have the latest ignite related patches.

live free or die
harry
Live Free or Die
Jerry Jaynes
Frequent Advisor

Re: Is this an ignite bug?

Ignite requires that you use a "no rewind" device, hence 0mn, not 0m.

Jerry
Never let 'em see you choke!!
Shannon Petry
Honored Contributor

Re: Is this an ignite bug?

I have not made images this way in about a year, so please correct me if I am wrong.

IgniteUX require(d/s) a dds1 driver as well the no rewind driver to make a bootable tape.
To configure the tape, use SAM.

I know I made a booboo 1 time, and had a system blow out only to find out it's image was put on a DDS2 tape. I now keep data on a cluster and if clients blow up I use the network install feature of Ignite as it's just as fast and easy to update.

Regards,
Shannon
Microsoft. When do you want a virus today?
Darrell Allen
Honored Contributor

Re: Is this an ignite bug?

Hi George,

I suggest you be sure you have the latest version of Ignite and use make_tape_recovery instead of make_recovery.

Earlier versions of make_recovery allowed you to specify a rewind device name for the tape drive even though documentation said to use the no-rewind device. I'm not sure of the reliablity of a make_recovery tape made using the rewind device. I have, in the past, successfully used them to rebuild a system however, I almost always rewind tapes before unloading and after loading them.

The current version of make_tape_recovery will error out if you specify a rewind device. I don't remember if earlier versions of make_tape_recovery did so or not.

Considering your error: It appears that make_recovery requires a tape be available even for preview mode (note that make_tape_recovery does not require the tape for preview). Was a tape loaded? Was it write-protected? Is it possible the tape was bad?

Just another reason to use make_tape_recovery.

Darrell
"What, Me Worry?" - Alfred E. Neuman (Mad Magazine)