Re: Handling unknown elements in an entry

["S. Mike Dierken" <mdierken@xxxxxxxxxxx>]

> > Why not? I do this for the back end servers at my
> > company (not for blogs and
> > not for arbitrary xml yet, and it's for a different
> > problem domain) & would
> > like to do more of it. It seems pretty useful as a
> > service. The storage
> > requirements and economics may be very different
> > between our two systems
> > though.
>
> I doubt you have millions of people using your back
> end storage at your company so our storage
> requirements and economics definitely are different.

Here's an example of data from some storage systems at my company:
http://www.amazon.com/exec/obidos/ASIN/1568812124

Although we have millions of people updating information, it's not
necessarily the same storage situation (fewer people continually adding
large blocks of text) or economic situation (we make money) as blogs. I do
realize these situations are different, but I was curious what kind of
numbers would be involved.

10M blogging users * 1K blog entries * 10KB per entry = 100TB of storage
Assume $100K per TB of high quality storage, redundancy, data center space,
etc. (this is a total random guess, not based on my work experience - I'd
love to hear what other folks think is reasonable). This is around $10M per
year or whatever.

What kind of rough calculations do you figure for your system?