Networker 7.6.1 - Windows 2008;
Last week one of our Backup at a site got messed-up – all of a sudden its
NAS Share become in-accessible – this Share had about 5 TB backup data,
Even after a few reboots and such the NAS Device/Share was still in-
accessible; Something went really NUTs on that Device/Share – at this
point it appears CBR (corrupt beyond repair).
This NAS Device/Share (source/site) is replicated to other Device (Target)
at a regional hub using Device's replication setup. Because Replication
promptly copies data-over, my hunch at this point is - this corrupt data
proporgated/replicated-over to the Target device share ( < at > regional HUB);
Now both copies of data are not much of use;
Luckily we do also have it Cloned to local tape – so not much problem
here ;
Now coming back to replication question – (because its mirrored via
replication, so if source loses data, so does Target) – This is because
the source D2D went NUTs!
In another possible scenario – where an virus or rogue intruder deletes
data on source (site) NAS Share (we have it strong password protection –
but we know how some of them smart/bad folks at cracking passwords or
punching holes); unfortunate side affect of blind Replication is – it
promptly mirrors all changes (including file delettions) from source
(site) device back over to regional HUB Device!! Thus data loss on source
propogates/replicates to the Destination - soon after!
Given this scenario, to prevent data-loss, few ideas come to mind:
1) Robocopy: At Target (regional HUB) - robocopy data from replication
share to some other share - once a week (may be keep 2-3 weeks
rotation/retention)
On Regional HUB's one server - put this in Task scheduler to run every
Sunday evening (by that time all regional replications finish):
ROBOCOPY \\HUBDevice1\site1_share \\HUBDevice1
\site1_share_backup /maxage:8 /S
ROBOCOPY \\HUBDevice1\site1_share \\HUBDevice1
\site1_share_backup /minage:21 /purge <-- does this work as intended ?
(ie, purge non-matching items - which are older than 21 days)
2) Networker backup of this backup-data, with 2-3 weeks of retention
(needs extra NAS-Share licenses)
It will be great if the Replication would have a feature:
A) Delayed delete feature in Replication (ie., Deletes at source will be
on HOLD for say 7 days or 2 weeks etc)
B) Anyway for Regional Device to automatically copy/replicate a NAS-share,
to auto copy one of the Share (Target) to some other local Share on Device
itself ?
We would like to perform this operation once a week or so – so even if
Source device at site lost its data data, at least we would have one week
old data to work with ?
As always - your insights and wonderful thoughts do count - have a
Valentine flower-bunch to y'all in advance!
Thanks
Sri
via RSS at http://listserv.temple.edu/cgi-bin/wa?RSS&L=NETWORKER
