Thanks for the input.
Since the ssretent is greater than the clretent, the clone savesets are
still fully browseable and recoverable even after the clretent
"expires". The savesets won't be recycled until the ssretent time is
passed. This is why I was thinking to force a new ssretent on the clone
ssids to make it easier to find tapes coming back. Otherwise, I need to
include logic on determining expiration based on clretent, instead of
making NW work for me and automatically recycling the clone ssids. Much
easier to query that.
It's a shame that we have to deal with this at all. Most other backup
products have built in DR management. My main goal, since I am forced to
create a DR process around cloning, is to keep things as simple as
possible.
The reason I picked 3 days is because this is for DR. You need the last
full and the last incremental (L9) to build a system. With L9s, I don't
have to keep several days to build the system. I figured 3 days to
guarantee I have something offsite - one tape going off, one in the
vault, and one coming back at all times - same philosophy with the
fulls, but they would be 3 weeks since they are taken weekly.
We have limited resources, so I was trying to take a conservative
approach by first splitting up test and prod, and then going with these
offsite rotations that I explained. I am not certain this is the best
approach, because of what I don't know about NW. Conceptually it seems
OK to me.
Any additional comments are welcome. This is a very subjective topic you
can't run by support, and there are no local EMC legato experts to go
over DR/cloning concepts without being charged an arm and a leg. All we
have is this listserv and any peer relationships you may be lucky enough
to have.
Many thanks.
-----Original Message-----
From: Francis Swasey [mailto:Frank.Swasey AT uvm DOT edu]
Sent: Tuesday, April 15, 2008 8:04 AM
To: EMC NetWorker discussion; Bruce Breidall
Subject: Re: [Networker] Cloning question
On 4/14/08 4:13 PM, Bruce Breidall wrote:
Assume that I manually cloned a saveset that has a browse and
retention of
91 days, but the clone pool has a retention of 3 days. If I cloned
that
savest today, the ssretent would be today +91, and the clretent would
be
today+3.
When it is time to figure out which volumes to pick to come back
onsite,
and you used mminfo to determine which volumes in the clone pool
have "expired" in terms of the clretent", could you then just relable
that
tape into your SCRATCH pool and mark it recyclable without messing
with
the original savesets that have a much greater retention? Would it be
better to set an actual ssretent on the cloned savesets via the
nsrclone
command, in addition to specifying the retention time on the clone
pool?
I believe NetWorker will figure out which ones have expired based on the
clretent and you will be fine with relabeling those tapes into your
SCRATCH pool. If you change the ssretent, you change it for ALL clones
of the ssid. Your description sounds like you don't want to do that.
What I am trying to do is separate clients by prod and test groups and
pools, and manually clone only the prod data. Since we do fulls once
per
week, and L9s the other 6 days, I wanted to try and have a three week
offsite rotation for the fulls and a 3 day rotation on the incrs.
If you only want the incrs for 3 days -- why bother to clone them at
all?
Does this sound feasible to anyone? Thanks.
To sign off this list, send email to listserv AT listserv.temple DOT edu and
type "signoff networker" in the body of the email. Please write to
networker-request AT listserv.temple DOT edu if you have any problems with this
list. You can access the archives at
http://listserv.temple.edu/archives/networker.html or
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
