Archives of the TeradataForum

Message Posted: Fri, 03 Mar 2000 @ 07:54:56 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Steve,

Option 1 and 2 are the same, really. In your script, you specify two FILEnames (that is, 2 volumesets). In the ASF2 GUI, you click on the "mirror" checkbox; on the command-line, you specify that the job is indeed a mirror job (-m).

This is what the script looks like:

     ARCHIVE DATA TABLES <...>,
     FILE = , FILE = ;

This is the command-line:

     /usr/XASF/bin/arcmain -s -g0 -tDLT -d2 -D, -Proot/asf2 -R, -C1 -m ......

The statement "The ASF2 mirrored option does not do parallel I/Os to the two tape drives." means to say that there is a performance penalty when doing mirror backups.

Important note: when doing ASF2 mirror backups, make sure that both drives are directly SCSI-attached to the node running the archive, or you'll end up sending all data for one the archives across a network, and that might not be what you want (unless it is a high-speed dedicated network). I was involved in setting this up at a customer site, and it is working well for them (given their volumes, backup- window and hardware).

I doubt that option 3 will work out, as it is REELBackup specific. I have not been able to find a REELLibrarian equivalent.

I know that some sites use option 4, strange as it might seem.

Your ideal solution -to the best of my knowledge- does not exist, at least not on the UNIX side. For people doing mainframe backups, there is a mainframe utility that does this.

Hope this helps,

Luc VAN GENDT
Sr. Customer Product Support Engineer / HAS

	<Prev	Next>		<<First	<Prev	Next>	Last>>