Data Security and Backups

Monday 12 March 2012

Managing Ageing Data

I had this discussion with several of my clients stating they have old data of hundreds of GB’s on their servers and they hardly require them once or twice a year or even not at all. Some of them state they have written DVD/CD’s of that data as backup, but that data still exists on Servers which they cannot delete as DVD/CD's are prone to failure. Those data is occupying a whole bunch of space on their servers which they can utilize in storing their current data. Don’t get me wrong those data are critically important data but storing everything on a single server doesn’t make sense. Adding more hard disk to the servers is definitely one of the options available, but I would rather implement a different system
Find out age of your data. There can be
a)      Young data
b)      Old data

Young data is the one which is highly valued and may be required to access almost everyday or fortnight.

Old data is the one which is critical but may not be required in daily functioning i.e. data that is year or 2 years or more older.

For e.g. Accounting data which is current and previous year can fall in category of Young data.
Data which is older than 2years can be considered as Old data.
We are supposed to maintain those data for taxation and government rules. Next step is how to get a system which takes care of both Young and Old data.

The Solution

As we saw there are multiple tiers of data.
Young Data can be stored on higher performance servers or devices.
Old Data can be moved to less expensive or slow performing servers / devices.

The following shall be benefits of this solution



ü  Save Costs
ü  Gives faster access to young data
ü  Better security both for young and old data rather than relying on
    DVD/CD’s or any other failure prone devices.
ü  Access to both of your data anytime you require.
ü  Reduce backup time as Old data does not change frequently so you
    can take backup of the same once a month or once every quarter.

Maybe currently you are treating all data equally and storing them all on one data. Imagine, if a disaster occurs on your server, you might loose everything. Again, we increase overload on the server by storing everything on that same machine.
We have to learn to implement a system which shall differentiate between ageing of data.

No comments:

Post a Comment