Storage I/O trends

Is more always better? Depends on what you are doing

As with many things it depends, however how about some of these? Is more better for example (among others):

  •   Facebook likes
  • Twitter followers or tweets (I'm @storageio btw)
  • Google+ likes, follows and hangouts
  • More smart phone apps
  • LinkedIn connections
  • People in your circle or community
  • Photos or images per post or article
  • People working with or for you
  • Partners vs. doing more with those you have
  • People you are working for or with
  • Posts or longer posts with more in them
  • IOPs or SSD and storage performance
  • Domains under management and supported
  • GB/TB/PB/EB supported or under management
  • Mart-time jobs or a better full-time opportunity
  • Metrics vs. those that matter with context
  • Programmers to get job done (aka mythical man month)
  • Lines of code per cost vs. more reliable and tested code per cost
  • For free items and time spent managing them vs. more productivity for a nominal fee
  • Meetings for planning on what to do vs. streamline and being more productive
  • More sponsors or advertisers or underwriters vs. fewer yet more effective ones
  • Space in your booth or stand at a trade show or conference vs. using what you have more effectively
  • Copies of the same data vs. fewer yet more unique (not full though) copies of information
  • Patents in your portfolio vs. more technology and solutions being delivered
  • Processors, sockets, cores, threads vs. using them more effectively
  • Ports and protocols vs. using them more effectively

   Storage I/O trends

Thus does more resources matter, or making more effective use of them?

For example more ports, protocols, processors, cores, sockets, threads, memory, cache, drives, bandwidth, people among other things is not always better, particular if those resources are not being used effectively.


Likewise don’t confuse effective with efficient often assumed to mean amount of something used.


For example a cache or memory may be 100% used (what some call efficient) yet only providing a 35% effective benefit (cache hit or miss) vs. cache turn (misses etc).


Throwing more processing power in terms of clock speed, or cores is one thing, kind of like throwing more server blades at a software problem vs. using those cores and sockets not to mention    threads more effectively.


Good software will run better on fast hardware while enabling more to be done with the same or less.


Thus with better software or tools, more work can be done in an effective way leveraging those resources vs. simply throwing or applying more at the situation.


Hopefully you get the point, so no need to do more with this post (for now), if not, stay tuned and pay more attention around you.


Ok, nuff said, I need to go get more work done now.


Cheers  Gs