For some reason, there are many businesses that believe their data should be secured and stored forever, after all, someday they might need it!. It really doesn't matter what your company policies and procedures were for keeping all your data from back in the 90s, any reasons have become mute in this day and age. The data you are so desperately trying to keep is obsolete, outdated, and of no value. Any information that was possibly in old data has probably been recreated in new data, so clean house. If you are someone who loves to collect and store images of animals, whether cats, dogs, horses, tigers, etc. You probably have every shape, size, and breed possible. The point is, after a few years, you are going to discover this data holds no value, so why keep it?
Artificial intelligence is not going to provide you any interesting information about your old company policies and procedures for keeping outdated data that you are holding on to. This old data won't even be pulled up by your search tool unless you were to literally search by “inventory retention procedures for 1990”. Even if you have a data storage company, it is not going to grab absolutely every single step you have taken since the beginning of time. What once was your Java EE applications server, is now Node js, so traveling back in time is not going to be of any significance, it's just taking up space.
Some people will keep data in hopes that one day AI will find something useful among this stuff. Most machine learning falls into regression, classification, and clustering. Clustering only groups stuff together that seems similar but chances are slim to nil that your old logs back in 2006 will hold anything useful or valuable to hang on to. If you really have the time to waste, you could always train your machine learning to find something useful in old stuff, but most of us don't.
Back to storage, even though storage itself is cheap, insight and organization are not. So what, you got a great deal on SAN or have been operating some mirrored JBOD setup with a clustered file system. The point is, that does not mean that storing noise is going to be cheap. Take into consideration the cost for the manpower for organizing, maintaining, and keeping your old stuff! In this day and age, search technologies are good at sifting through relevant material vs what's not, but it does come with a price tag. All said a done, spring is just around the corner so it's time to start performing some significant Spring Cleaning!
Artificial intelligence is not going to provide you any interesting information about your old company policies and procedures for keeping outdated data that you are holding on to. This old data won't even be pulled up by your search tool unless you were to literally search by “inventory retention procedures for 1990”. Even if you have a data storage company, it is not going to grab absolutely every single step you have taken since the beginning of time. What once was your Java EE applications server, is now Node js, so traveling back in time is not going to be of any significance, it's just taking up space.
Some people will keep data in hopes that one day AI will find something useful among this stuff. Most machine learning falls into regression, classification, and clustering. Clustering only groups stuff together that seems similar but chances are slim to nil that your old logs back in 2006 will hold anything useful or valuable to hang on to. If you really have the time to waste, you could always train your machine learning to find something useful in old stuff, but most of us don't.
Back to storage, even though storage itself is cheap, insight and organization are not. So what, you got a great deal on SAN or have been operating some mirrored JBOD setup with a clustered file system. The point is, that does not mean that storing noise is going to be cheap. Take into consideration the cost for the manpower for organizing, maintaining, and keeping your old stuff! In this day and age, search technologies are good at sifting through relevant material vs what's not, but it does come with a price tag. All said a done, spring is just around the corner so it's time to start performing some significant Spring Cleaning!
Why You Should Not Keep All Your Data Forever
Reviewed by thanhcongabc
on
September 21, 2017
Rating:
No comments: