rrestore crashes tape host
Thanks to all who replied with many wonderful suggestions.
I also received enough "me too, what gives" messages to justify
a general summary posting. My original message:
> After upgrading to SunOS 4.1.1, every attempt to perform
>an interactive restore, from a multi-volume 8mm tape, crashes
>the tape host with the following message:
>Apr 15 17:59:35 namu vmunix: st0: tape synchronization lost
>Apr 15 17:59:35 namu vmunix: st0: file positioning error
>Apr 15 17:59:35 namu vmunix: panic: psig action
> The command that causes the crash (from multiple remote hosts):
> restore -ivfs namu:/dev/rst0 #
> The platform is a 3/50, running 4.1.1a on a Maxtor XT-8760S
>with a separate 8mm tape unit. The system is running the
>GENERIC kernel as distributed with 4.1.1. The configuration
>performed flawlessly under SunOS 4.1. Remote dumps are still
>working without failure and local restores (full and interactive)
>continue to work.
> Anyone else tripped over this yet?
The solution: install patch number 100211-02 to fix the tape
synchronization problem existing in 4.1.1 (and supposedly in
4.1 as well). This involves installing a new st.o in your kernel.
Other suggestions included installing the patch to resolve psig
crashes discovered on exit from sunview. I'm a firm believer in
installing only as many patches as necessary, so I can't comment
on that approach. Solving the tape sync problem has eliminated
the tape host crashes, thanks!
Many replies suggested abandoning the -s option to restore and
manually forwarding the tape with mt. I would have considered this
option as a work-around if the available patches had not solved the
problem. All of the utilities had functioned properly prior to the
upgrade and have been installed on several tape hosts in the dept.
Installing the patch when necessary was the best solution for us.
Bryan Rawson
School of Oceanography WB-10 E-Mail: rawson@ocean.washington.edu
University of Washington Phone: (206) 543-0676
Seattle WA 98195 FAX (206) 685-3354
