Help! Read DLT-IV tape on SDLT drive fails

[jruslan]

We are moving our NT4 Legato Backup Server to UNIX based Solaris 8 backup server. I have some trick on howto move indexes from NT-2-UNIX. AND.. I need to use the tape from old NT4 server with DLT7000 devices and scan the tape on the new SDLT220 device (mmrecov is important). When I start mmrecov on SDLT I have the following error:
======
What is the name of the device you plan on using [/dev/rmt/1hbn]? /dev/rmt/8hbn
Enter the latest bootstrap save set id: 3064065537
......
Scanning /dev/rmt/8hbn for save set 3064065537; this may take a while...
scanner: error, read open error, I/O error
uasm: invalid save stream
/dev/rmt/8hbn: Mount operation in progress
/dev/rmt/8hbn: verifying label, moving backward 2 file(s)
mmrecov: RAP error: sdlt tape AT0178 not in media index
mmrecov:
Cannot Mount volume on /dev/rmt/8hbn
mmrecov: uasm exited with error 1
mmrecov aborted.
======

But in my previous tests I've used DDS4 and that works just fine. So, could you please advice if there are errors in st.conf or I need to change "block size" on UNIX side, etc.

P.S. Whats appropriate st.conf line for SDLT? Mine is here:
SDLT-data = 1,0x38,0,0x8639,4,0x90,0x91,0x90,0x91,3;

JR

 

[Zantor]

I would check the blocksize first. The default for NT4 is 64 KB.

 

[jruslan]

I did it, block size has been changed and networker shows that changes "block size now 64k" or something like that. But it doesn't help at all
Here is the way I changed block size (this is "csh&quot:
#setenv NSR_BLOCK_SIZESDLT 64

 

[Riaan]

this kind of cross-platform migration is not supported by NetWorker. There are all kinds of issues with big-endian vs little-endian hardware platforms. On the NetWorker mailing list (

http://listmail.temple.edu/archives/networker.html)

there were even guys who were unable to move NetWorker from one RISC Unix (e.g. HP-UX) to another (e.g. HP Tru64)


Riaan van Niekerk


UNIX sysadmin

Potchefstroom University

South Africa