Re: [Networker] Working with ADV_FILE devices
2005-08-17 14:07:27
Ronquillo, Merill CONT (NAVFAC) wrote:
Just trying to understand more about adv_file type devices. Suppose I have a
pool of 4 adv_file devices. Server parallelism is 64, target sessions on each
disk device is 8. If I understand target sessions correctly, this is how the
data gets backed up:
Volume1 gets savesets 1-8
Volume2 gets savesets 9-16
Volume3 gets savesets 17-24
Volume4 gets savesets 25-32
Volume1 gets savesets 33-40, THIS IS WRONG (even for tape devices)
Volume1 gets SS 33
Volume2 gets SS 34
Volume3 gets SS 35
Volume4 gets SS 36
etc.
What would happen if savesets 1-8 fill up Volume1 and there's space remaining
on Volumes 2-4?
Save sets 1-8 will wait until space on Volume1 is reclaimed by staging
or deletion. Save sets to adv_file devices do NOT span between volumes.
Ever.
Would savesets 33-40 be backed up to Volume1 after data already residing
on the volume is automatically staged
SS 33-40 will go to whatever devices are available at the time, although
if a device is full and waiting for space, no new save sets will be sent
to that device.
Or will Networker see that Volume2 has free space and send 33-40 to that
volume?
It is very unlikely to send all these sessions to the same device, it
will spread them around based on current sessions and target sessions.
The amount of space remaining is NOT what governs where the save sets
will go, except possibly in the case where all devices have exactly the
same number of sessions saving to them.
This is all a little academic, since multiple adv_file devices in the
same pool is not the recommended way of setting things up.
--
Note: To sign off this list, send a "signoff networker" command via email
to listserv AT listserv.temple DOT edu or visit the list's Web site at
http://listserv.temple.edu/archives/networker.html where you can
also view and post messages to the list. Questions regarding this list
should be sent to stan AT temple DOT edu
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|
|
|