Thanks to Tom Blinn for his answer.
It seems I have bad luck. This product, though very useful, is not very
commonly used and therefor hardly tested against the patches. I did some
debugging and found an error: the tr-commands, used in the btcreate and
mktape scripts did not translate as one should expext. I changed them from
'tr "[a-z]" "[A-Z]"' into 'tr "[:lower:]" "[:upper:]"' which at least gave
better results, but didn't solve the problem. I will do some more testing an
report it here.
Tom Blinn said:
Do you have a Compaq support contract? If so, report the problem to your
support center and request a fix. If not, you are probably out of luck.
Relatively few people use the bootable tape facility; it was put into the
product and supported to meet the needs of a specific OEM partner, and I
suspect the bulk of the testing (if any) is oriented toward assuring the
systems they require work as they configure them. I am almost certain
that it is NOT something that is verified with patch releases; that is,
we could break bootable tape in a patch release and not know it until we
got a formal complaint from a supported customer. (This is not the way
things SHOULD work, but it's the reality of the situation.)
Good luck. You are probably going to need it. If you are planning to
depend on using the bootable tape facility along with patches, then you
probably need to plan on verifying that the combination works with each
new patch release on your configurations before you deploy the patches
in your environment, because that is NOT being done by Compaq's patch
release test team.
Tom
Dr. Thomas P. Blinn + UNIX Software Group + Compaq Computer Corporation
-----Ursprüngliche Nachricht-----
Von: hans van der heide [mailto:hans.vanderheide_at_bluewin.ch]
Gesendet: Montag, 3. Juli 2000 12:45
An: 'tru64-unix-managers_at_ornl.gov'
Betreff: btcreate / btextract hangs when booting from tape
Hi folks,
I have the following problem:
When I boot from a tape, created wit btcreate, with the command 'boot -fl
"nc" MKA500' the system will read in the bootblock and vmunix, but then
hangs forever after showing it will work in lazy swap mode. Obviously it
does not want to start btextract.
This occurs on different system types. In the past (Tru64 Unix 4.0D patch 3)
I could use this procedure successfully, but since I installed Patchkit 6,
it does not work any more. I tried to use the old btcreate/btextract/mktape
combination which worked for me before, but that did not work as well. It
seems some AdvFs patches interfere with btcreate/btextract. Luckily I can
boot from CD, skip the first 2 files, and then vrestore the saved files by
hand, so I don't have to fear I cannot restore the system. But I'd like to
have the nice boot-facility back for my +100 systems backed up in this way.
I have seen this problem mentioned before, but did not read any
solution/bypass.
Data:
O/S Tru64Unix 4.0D Patch 6,
System Disk: Root and /usr are both AdvFs with enough free space:
Filesystem 1024-blocks Used Available Capacity
Mounted on
root_domain#root 131072 56906 68336 46% /
/proc 0 0
0 100% /proc
usr_domain#usr 716800 513381 180808 74%
/usr
Can somebody shed me some light?
Hans van der Heide (hans.vanderheide_at_bdl.ch)
BDL Informatik GmbH (www.bdl.ch)
Tel: +41 (0) 56 442'40'31
Fax: +41 (0) 56 442'40'33
Received on Fri Jul 07 2000 - 06:05:50 NZST