Discussing all things virtualization and storage in the data center.

USP V features

Now that the wild excitement of the release is over, I thought it would be worthwhile discussing some of the new features that I think have been included in the V.

The list on the web site is as follows?

<snippet>

  • Business continuity services
  • Content management services (search, indexing)
  • Non-disruptive data migration
  • Volume management across heterogeneous storage arrays
  • Thin provisioning
  • Security services (immutability, logging, auditing, data shredding)
  • Data de-duplication
  • I/O load balancing
  • Data classification
  • File management services

 </snippet> 

So, there are the normal ones but a couple appear to be new to me at least and if not, I need to pay more attention to what I have already.

Content management services.  I am not sure what they mean by that but if it is similar to something I once saw and dreamed of getting, it has to make for outstanding performance in "specified" data types.  Eg, you classify data into types, such as XML files, voice captures, emails, jpeg files, etc.  A solid state database holds the location of the specific data type and file name and knows exactly where it is on disk.  So, there is no searching ever.  I have seen databases built to do the same thing where when you ask for a file, the database has a pointer to its location but it is still a slow process when you have millions of them.

Thin provisioning.. lets see about this.  Will it work on UNIX file systems because of the super block?  Can it be used for replication?  In other words, can a dynamically allocated CU:LDEV grow at the replicated end automagically?

I am still trying to figure out what immutability has to do with Security Services.  Perhaps the word has a new meaning other than changeless or something to do with Charles Darwin?

Data de-duplication in particular is interesting.  After seeing what Riverbed can do to improve (in particular) WAN performance by de-duplication of data, it makes me wonder if the V can do something similar.  OK, some data will never be able to be de-duped but if we can save lots of money by freeing up disk space, that's got to be good right?

The last two seem to be related to what I said about Content Management well, at least in my wish list.

I am eagerly waiting to find out exactly what all these options can do for me and if EMC can do it better.. just to be fair on them.

Stephen 

comments powered by Disqus