Hi!
I have a problem with a slow and confusing "Client-Index based" recover..

My enviroment:
I have a win2k server with 2x933 Mhz cpu, 512 Ram and NW 6.1.3.
My client is a Sun sunfire 880 Solaris 5.8 with NW 6.1.3..4 x sparccpu 8 GB
RAM.. it has mounted disks that are my savesets:

/fakt15
/fakt16
/fakt15/FF
/fakt15/DYE
/fakt16/IPR
/fakt16/SYST
/fakt16/TSYS
/fakt16/ASST
/fakt16/DUTFT

I do a "recover -i Y -d /newdest -s backupserver"
adding my /fakt15 savset.
the list is prepared and it is 78686 files..
I start the recover and after 10 minutes the recover list is sent to
nsrmmd? and are then 93996 files!
Very strange..

Then i'm starting a new recover session and adding my other saveset /fakt16.
It is 194519 files.. I start it and it takes so long as 50 minutes before
it starts.. Then it request 194818 files.

So the questions..

1. How can I speed up the recover process so it starts faster? is the
backupserver (with nsrindex and nsrd) the bottleneck? Should I have a linux
or unix server instead..

2. Why are the amount of files different under the recover process.. I cant
understand it..
I add /fakt15 and its 78' files.. The process adds all underlying catalogs
and savesets /fakt15/FF/fakt15/DYE etc and it are then 93' files. When Im
adding /fakt15 and all other underlying saveset they are togehter is 90500
files..which are the other 3000 files?

Any ideas and suggestions?

Kind regards and thanks in advance
Fred

--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listmail.temple DOT edu or visit the list's Web site at
http://listmail.temple.edu/archives/networker.html where you can
also view and post messages to the list.
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=