IP backup solutions?

[bscott at ntisys.com]

On 14 Sep 2003, at 11:45am, lists at karas.net wrote:
>>> ... optimized for sending lots of data (many gigs) over a slow pipe (T1
>>> in this case) ...
>> 
>>   I suspect that is not reasonably possible.
> 
> Actually, it *is*, but not in a very desirable manner.  It seems that most
> of the IP/remote backup solutions use a lot of proprietary compression,
> and schemes to back up, literally, only the bits (not files) that have
> changed in some cases.

  Well, if the diffs are not gigabytes, then yes, it may be feasible.  But
you didn't say that; you just said "many gigs".  :)  Do you have information
on what your actual data change patterns will be?  What are we actually
backing up here?  Keep in mind that the extreme cases count; you don't want
to end up saturating the line for a week because someone dumped 10 gigs of
new data onto the network in one day.

  As others have suggested, rsync can keep two filesets synchronized, using
just the diffs.  You could then backup the fileset at the main office.  But
again, that depends on the size of your regular diffs.

>> I personally find the whole situation contains far too many unknowns.  I
>> generally require backups to be dependable.  Taking chances just isn't
>> worth it.
> 
> Yes, it does, but when you get customers and sales people involved without
> enough tech consultation...

  Same thing that happens when you get the wrong people doing surgery or
building airplanes.  Very Bad Things.

>> Obtain a decent tape drive (for small offices, DDS4 is very
>> cost-effective).
> 
> Excellent suggestion, and probably the "ideal" solution, but not feasible
> for this project.

  Again: If you think it costs a lot to do it right, just wait until you
find out how much it costs to do it wrong.

  If this is a case of "You've told people that but they won't listen", I
recommend making sure your objections, and the decisions of the people
overriding you, are in writing.  You don't want to be the one taking the
blame when the excrement inevitably hits the air circulating device.

-- 
Ben Scott <bscott at ntisys.com>
| The opinions expressed in this message are those of the author and do  |
| not represent the views or policy of any other person or organization. |
| All information is provided without warranty of any kind.              |