The Bob's Honest Truth Bob Sneed's Blog

"Honey, I just heard on the radio that Oracle just bought Sun."   Those were the words my lovely and talented wife used to wake me on the morning of Monday, April 20th, 2009.  "Hmmm", I thought, "I guess I should start my day by peeking to see what's on the wire."  After unhurried but minimal morning ablutions (the norm for many work-from-home IT workers), I asked The Great Google: "(oracle sun)?" "You are #3!", spake The Google.

Cool!  My 2001 Sun Blueprint "Sun/Oracle Best Practices" was the third hit!   Quickly, I took a snapshot of that window for posterity!   Musing in mixed metaphors, I thought "Gee, The Prisoner was only Number 6!"  In the minutes that followed, my Blueprint quickly plummeted from its #3 spot, rapidly displaced by gaggles of news and speculation.  "Well!  That was affirming!", I thought;  "Maybe I should revive my blog?"

I stopped blogging shortly after I started in 2007 because of my health issues with degenerative disk disease, nerve impingement, arthritis, and chronic pain - all of which adds considerably to my baseline challenges of being ADHD to begin with!   However, I have gained some ground on that front in the last year, and the new doctor with whom I consulted Wednesday gave me some renewed hope as we laid out a new treatment strategy.  I'm still keyboard-limited, but as-ever bulging with choice tidbits of knowledge regarding Sun/Oracle performance and capacity topics - straight from the front lines of Sun/Oracle pre-sales and escalation support.  I reckon it's a good time to try getting back on the blogging horse!

What should I write next?  What do you think?  I have numerous blog-size articles hanging around in the concept or draft phases, including these:

• "Oracle I/O Eye Chart" - A modernized version of my old "eye chart" slide summarizing filesystem features for Oracle deployment; perhaps adding the dimension of "contiguity".
  
• "How to Succeed (or Fail) with Oracle on UFS" - A how-to guide.
  
• "How to Succeed (or Fail) with Oracle on QFS" - A how-to guide.
• "How to Succeed (or Fail) with Oracle on ZFS" - A discussion of implementation guidelines, motivations, and tradeoffs.
  
• "How to Succeed (or Fail) with Oracle on SPARC CMT Systems" - A totally descriptive title, I think, eh?  This blog entry would merely be an introduction to the slide deck I'm currently working on with Glenn Fawcett.  We will be seeking to present it at some conference, somewhere, sometime soon.
  
• "Eschew Direct Obfuscation" - More illumination on the surprisingly-non-obvious-and-diverse dimensions of 'direct I/O' with Oracle on Solaris.
  
• "From 80% to Chaos" - Maybe that's not a 'hockey-stick' curve you are looking at?  Maybe it's the gods trying to signal 'S' for 'Stupid' on their Ouiji boards?  I might be able to expound on this from here to eternity!  Alternately, "It's 80%; does Solaris know what your priorities are?"
  
• "ISM, DISM, Drazzle, Drone" - A compendium of 'memory QoS' topics pertinent to Oracle performance on Solaris.  "Help, Mr. Wizard!"
  
• "Waiting for I/O" - Putting "wait i/o" to bed ... forever!
  
• "Oracle cpu_count in Modern Times" - Considerations for Oracle cpu_count setting on modern SMP, VMT, CMT and virtual systems; culminating in some recommendations for various deployment scenarios and proposals for changing the default in future Oracle releases.
• "Adding Concurrency to a Long-Running Oracle Spatial Application" - A happy success story, with sample C code.
  

Article concepts for which I am still accruing knowledge include the following ... which will therefore not be 'next'!   I wonder what's the interest level in these relative to those above?

• "How to Succeed (or Fail) with Oracle on Sun Open Storage" - Performance management takes on new dimensions with ZFS baked into an affordable Fully Integrated Software and Hardware storage solution that leverages flash-based SSD technology and ground-breaking user-accessible I/O analytical capabilities.
• "Adventures in Capacity Planning" - Revelations into the often-tragic and unsatisfying state-of-the-practice.