[comp.sys.apollo] RBAK problem rbak'ing from sun.

rmallett@ccs.carleton.ca (Rick Mallett) (05/16/91)

I am attempting to adapt our Sun backup scripts to handle Apollo
workstations and I ran into the following problem. When I sit on 
an Apollo and execute the command:

rsh sun "dd if=/dev/nrst0 bs=8192" | /com/rbak -stdin -index -all

it works, but when I try to initiate the process from the sun 

dd if=/dev/nrst0 bs=8192 | rsh apollo /com/rbak -stdin -all -index

I get the following error. 

?(rbak) Could not dup stdin.$?uld not open backup file. \
   - can't open stream to this object type (stream manager/IOS)

Is there any way around this? I've tried a million variations, but
I always get the same result. I assume it has something to do with
attempting to dup stdin without a pseudo terminal attached to the 
session, but I don't have a clue how to get around the problem. I
need to initiate the command from the sun for operational reasons.
I was able to get it to work by cat'ing to a tmp file and then 
running rbak against the file, but I often won't have the disk 
space to do it that way. Any suggestions would be appreciated.

By the way, wbak'ing accross the net to an exabyte on the Sun
seems to work fine, and you can then rbak interactively using
the first command above, if that happens to be of interest.

Actually, what I would prefer to do is index the backup tapes on
the Sun instead of rsh'ing to the Apollo. Does anyone have any PD
source code which will scan an rbak volume and produce an index.
Better yet can anyone tell me the format of the tapes so that I can
integrate the indexing operation into my Perl scripts. I don't mind
rsh'ing to the apollo for restore operations since they happen 
infrequently. Also, I know that I can get wbak to produce virtually
the same index listing, and I may have to go that route, but I 
like to read the tape to verify that the backup is good.