Data starts when the data is created. Need bigger backups, need more primary storage.
Storing is not the hard thing, the harder thing is managing the data.
90% of errors are human errors. So risk to manage data is as important as storing data
There is a decline in budget, but applications are making more and more data
Data should be classified, and sent to the right tier based on what type of info it is.
How long does it take to restore from tape? What happens if you have to freeze data based on a court order?
According to symantic – nearly half files backed up are over 90 days old.
Files over 90 days old occupy ~17% of toital storage (abt 4.8T)
From audience: files may be over 90 days but you still need them for a biz reason, dedupe and archiving helps with that
From audience: why does it have to be object storage?
So what’s more important? meta data? tiering? object storage?
It depends. 🙂 Scalability, non-classifiable data (x-rays),
Mail admins – psts and local machines. Worst case scenario for mail admins cuz you dont have access to pst on local laptop
DX based on rack-based x86 server platform
From audience: In small environment wouldn’t another EQL rack fix problem? It depends. DX may be the solution if you are in a regulatory situation
DX objects are ideally suited for unstructured data, just add nodes as you need them.
from audience: does object storage have enterprise vault installed on it? Answer: yes
American Systems reduced backup time by 65%
Formula 1 team had a 50% reduction in email storage and server infrastructure
Mitel saved 8million (US) in labor productivity gains, const savings, etc
Understand customer needs, and build based on numbers gathered, etc
Solutions are co-engineered and tested