Just noticed a very curious "feature" of NetWorker 7.2/7.1.3. While
setting up some test-bed systems, we were doing a Directed Recover from
Client A to Client X of about 75GB. At the same time, we initiated a
second Directed Recover from Client B to Client Y. I anticipated that the
second recover would simply be queued up until the first recover had
completed, but lo and behold, the second recover finished first (much
smaller dataset). I assumed that this was a fluke of timing, in that the
two datasets happened to be multiplexed onto the same tape in the same
session. Kind of cool.
But we then initiated a recover to Client Y of the same dataset that was
still being recovered to Client X. I assumed that as a best case, it would
start recovering the same files that Client X was currently recovering,
then backtrack and pick up the rest.
But instead, it "paused" the first recovery to Client X (45GB out of 75GB),
backtracked to the beginning of the saveset, started recovering to Client
Y, then when Client Y caught up, both clients continued in parallel!!
My question is - is this expected behaviour, or a fluke? Or is it an
unanticipated "feature" introduced to accommodate the multi read/write
features of Advanced File Devices (even tho these recoveries are from
tape)? How then are priorities determined for recoveries?
In this situation, it worked to our advantage; but what about the reverse?
If I'm trying to recover a small set of spreadsheets that my boss nuked by
accident, for example, and someone else kicks off a recovery of a 60GB
database, am I potentially going to have to wait for the other to finish?
Just wondering,
caf
--
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
=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=
|