Flash Cache at Chuy's

As Kong continues his job search we found a place that was hiring for this week's episode of  So Say SMEs. Chuy's is of course the perfect place to discuss the new vSphere Flash Read Cache feature aka vFRC.



I decided to wear a vintage work shirt this week - from a company that no longer exists.  Anybody remember Altiris?

Let The TPC-VMS Benchmark Wars Begin

In the benchmark world, the Transaction Processing Performance Council (aka TPC) has long been the owners of the premier performance benchmark.  I'm referring to the TPC-C database performance benchmark where millions have been spent to prove performance excellence.

A new benchmark from TPC that is designed to focus on the performance of virtualized databases was created called TPC-VMS.  And very recently the first TPC-VMS result was published by HP using vSphere 5.5.

This means that the HP and VMware vSphere solution has the "Leadership" position for the TPC-VMS benchmark - as of right now.

What happens next will be interesting.

The first published result establishes the target for everybody else.  Will there be a bunch of publications over the coming months?  Each a little bit better.  Or will it take some time for anybody to be able to surpass this initial publication?

It is also of course possible that nobody else will ever publish a TPC-VMS result.  Sometimes new benchmarks just never get established.  If this is the case, then vSphere would hold the number one position forever.  But this isn't what we want.

Ideally the healthy competition of a benchmark fosters innovation, software and hardware improve, costs come down, and customers benefit.

VMworld and Fantasy Football 2013

Bringing together VMworld and Fantasy football for 2013.  I also answer some of our critics with a bold statement.  We are not only about enterprise solutions and football.

ESXTOP in The Hands On Labs At VMworld 2013

I'm one of the captains for the performance module of the Hands on Labs at VMworld this year.  This means that I got to really geek out and create a new module that focuses on using esxtop.  For really in depth performance analysis on ESXi - esxtop is my favorite.  Come by the Hands On Labs in Moscone South this week take SDC-1304 Optimizing for Performance.  All of the modules are great - but my new esxtop module is number six.  Skip straight to it by using the table of contents once the lab has started.

The labs are currently only available onsite at VMworld, but will be made available through the public Hands on Labs portal at some point after the conference ends.  So if you couldn't make it to VMworld this year - don't despair, you will be able to get to the new labs in the future.

Software Defined Is Everywhere

Kong and I returned to So Say SMEs with this new episode about Software Defined EVERYTHING.  While the term is definitely a bit over-used right now, we talk about it anyway and use it a bit more, making it  even more over-used. Hopefully we also get the heart of what it actually means.



Kong did not send me the memo on wearing a cool Hawaiian shirt.  Maybe next time.

Do You Know Who I Am?

Just how long will it be before we get the mug shot photo?  We've all seen the celebrity mug shot after they are arrested for late night hijinx.

Once you hit the really big time your every move is followed closely.  There are twitter storms and blogs and even official news reports.  Unfortunately, it inevitably leads to an attempt to use fame to get around the norms and rules of society.

Imagine the scenario....

After a late night of hanging out at the Hands On Labs at a local VMUG, our celebrity guru blogger and author will get stopped on his way into the backstage area of the labs.  He just wants to grab a slice of the free pizza for the lab techs.  The security guy wants to see his badge.  The badge doesn't have the right color dot.

"Do You Know Who I Am?"

Unfortunately, this security guy has never installed ESX or vCenter or vCD.  He's never read about resource pools or pondered the best SRM strategy.  He just wants to see the red dot on the badge.

......

It even made this week's episode of So Say SMEs in Virtualization and Cloud!

Vote for Half of The Tong Show

In this week's So Say SMEs aka the Tong Show, we talked about VMworld 2013 Session Public Voting.



My sessions up for vote this year are 5591 Big Data: Virtualized SAP HANA Performance, Scalability and Best Practices and 5517 Performance In The Cloud Doesn't Matter.  Easiest way to find them is to just go to the voting site and filter on my last name - Muirhead.

Cloud Bracket Busters AWS and Hybrid Cloud

On this week's episode of So Say SMEs, Kong and I discuss the new announcement from VMware of a Hybrid Cloud service in the spirit of March Madness.


Notice the cool circuit board design on the wall behind us - anybody know where we filmed this episode?

Performance Perception In The Cloud

A funny thing seems to happen when applications are run in the "cloud" instead of just "virtual" in your local datacenter.  The perception of performance is changed.

As a part of my job, we look almost everyday at the performance of applications running in Virtual Machines.  We sometimes compare this performance against applications running natively on hardware and sometimes we are just measuring response time or looking for lost cycles.  Even small differences of less than a percent are important and improvements of whole percentages are celebrated.

Customers like to get into detailed discussions about how much "overhead" virtualization brings.  Settings and configuration options in vSphere that they can use to squeeze out every last drop are given in whitepapers, presentations, and blogs.  This is the kind of stuff that I consider fun and interesting.

But when we discuss running applications in the "cloud" the aspects of performance that are important change.  Even though all clouds are enabled by a virtualizaiton layer with virtual machines running around on top, the performance concern is almost always only about scaling.  Can additional instances be spun up to meet customer demand in real time?  Can storage capacity be expanded infinitely into the future?  There is absolutely no discussion about virtualization "overhead".

I believe that there are many reasons for this difference in perception.  Some of these reasons even make logical sense, but I don't think that the logical aspect is what is driving this difference in view.

It is primarily driven by the completely different approach that true "cloud" computing brings.  The underlying infrastructure is being turned over to somebody else, and the customer is simply purchasing a service.  While in the virtualization world, the underlying infrastructure is still under the control of the person making the decisions.  Even though both a cloud and a virtualized data center will be using the same virtualization approach, it is only the virtualized data center that is questioned about "overhead".

Interesting to me that the evaluation of performance is so different between these two.

Database Evolution Revolution

An excellent blog by Chuck Hollis today really got me thinking about databases.  In Chuck's post (on Chucks's Blog) he relates a presentation by Charles Fan that was recently given.  From Databases to Dataclouds is the title and a good overall summary of the blog / presentation.

After reading it my initial reaction is that this seems like a total revolution in how data is used and harnessed.  We are now talking about being able to use new database technologies and techniques to mange, control, and analyze data in near real-time.  Time-to-decision in seconds.  Data literally flowing all around us and yet captured, analyzed, and understood in real-time.

We would all like to see the world move on from relational databases to a totally new approach - a revolutionary approach.  But the reality is that the relational database will continue to be the key core aspect of just about any data intensive effort.  Additional features and optimization will continue to be added.  Entirely new ways to deal with and manage data can be used in conjunction.  But the core strengths of relational databases (transactions come to mind) will continue to make them indispensable.

I believe that what we are seeing is an evolution in database software combined with an evolution in hardware performance which adds up to a revolution when seen together.

When you take this and use it on top of the revolution that virtualization has created, you end up with a paradigm shifting view of the world.  Even though Chuck's blog is kinda long, I recommend that you read it through and see if it makes you start thinking too.

So Say SMEs: Little Things Can Be Big Later

In this week's episode of So Say SMEs, Kong and I discuss how small strategic additions can be big later.  We also manage to relate this to football - as usual.

A production note - the door behind us didn't open a single time during filming (unlike last week).  Although there were a couple of people that started to walk towards us right at the end, but finished before they were able to get on camera.

Super Bowl and Content Deduplication

On this week's episode of So Say SMEs, Kong and I discuss the end of the football season and content de-duplication.

If you watch you will see that the door behind us kept opening.  Normally we just roll with these types of events on So Say SMEs, but in the case the noise from the servers was too loud and we had to wait for the door to close again before continuing.  Awesome work by the team of video editors to put things right.

The Hitchhikers Guide To ESXTOP Counters

For all the other performance people out there that spends time using esxtop to do deep analysis of ESX performance, there is a great document that you really should know about.  It is called vSphere Monitoring and Performance for vSphere 5.1 / vCenter 5.1 / ESX 5.1.

My favorite section is a detailed list of all the esxtop counters and exactly what they mean and how they inter-relate.

For example, I had a question about what is the difference between the %USED and %RUN for a VM.  On page 155 and 156 of this great document it gives you the specific formula for %USED, which is:

%USED = %RUN + %SYS - %OVLP

It also explains in the %RUN definition that on systems with hyperthreading enabled, the %RUN value does not account for hyperthreading.  This can cause %RUN to be twice as large as %USED.  But wait this can't be possible based on the formula - except in this case it is because %USED accounts for hyperthreading.

Another mystery of the universe is solved with the help of the vSphere Monitoring and Performance guide.

Super Bowl Competition and Tech Cooperation

In this week's episode of So Say SMEs we have a fun talk about the Super Bowl and in true geek fashion manage to relate it to the tech world and then it back to Fantasy Football.

Included references to Hyper-V vs vSphere, Ravens vs 49ers, and Dell vs HP - so fun for everybody.

Attracting A Crowd at So Say SMEs

We had a lot fun last week with our season two kickoff video.  It was in a great location and we tried to set the stage for the year.

On the first day of the video being posted, I had a few retweets from others that I follow on twitter.  Which was really nice and much appreciated.  Then a few days later I got a whole slew of retweets from people I don't follow.  As you can see from the screen grab, there is a definite trend in the profile pictures.

Looks like my blog on the spammers getting it right might be true in this case as well.

More Cowbell and Stevie Ray Vaughan In So Say SMEs Season Two

Kong and I kicked off season two of So Say SMEs this week.  I promise more cowbell this year.
 

We did this initial video next to the Stevie Ray Vaughan statue near downtown Austin.  The reasons why are covered in the last minute or so of the video.  Here is the Stevie Ray song that I was talking about.

Apple and The Cult of Personality

There are a couple of facts about Apple today that may or may not be related.

• Charismatic and dynamic Apple founder and CEO Steve Jobs died a little over a year ago.
• Apple's stock has taken a dramatic downward turn over the past few months.

If Apple was really a cult of personality centered around Steve Jobs, then these two things are directly related. But I think this is an overly simplistic view of what is happening.

Apple's dramatic rise to become the most valuable company in the world was based on it creating entirely new markets with high margin products and then dominating those markets for many years afterwards even as competitors entered with less expensive options.  This was true with the iPod, iPhone, and iPad.

Without an entirely new market or category being created by Apple in the last couple of years, the  rate of growth has slowed.  While still dominant in it's existing markets, there are no new products on the horizon that would warrant the continued growth rate in the stock.

There is also the law of really big numbers at play with Apple.  It is has reached the point in size where it becomes very difficult to continue to grow at the same rate it did when it was smaller.  At some point growth slows.

It is possible that Steve Jobs would have been able to keep the growth up a little bit longer, but not too much longer.  This was no cult of personality, but it was and still is a great company.  It will continue on and while there are surely things that Steve would have done differently, it is difficult to believe that Apple would not have performed about the same.

SpamBot Gets It Write About Storage

The lengths that spammers will go through to try and get comments added to blogs has become pretty sophisticated over the years.

They started with just mass blanket efforts to put the same comment on every blog they could find.  These messages typically just contained a crass message and link to some not so credible website.

A more recent strategy was to cleverly write a generic message that could be applied to just about any blog post ever written, but to then contain a link to some not so credible website.  Something about how this is a very informative blog post and please continue to share you wisdom - link to random site that has nothing to do with anything.

They have finally progressed to the point where the spam is so targeted they actually are on topic and, at least in a recent case on my blog, topical and funny.

I recently wrote about the future of storage in a couple of blog posts, and a spammer blogged that the mini-storage units down the street were very nice and came highly recommended.  How could I have written a story about storage and not mentioned storage units?

Now if we can just figure out how to virtualize the mini storage place down the street, we can truly bring virtualization to the masses.

Hopping Over CNET

Everything that is written, produced, or created is influenced by the person or team that originates it.  In the press there are many very good people that do their best to report things in a fair and unbiased way, but even in deciding to write one story over another - a judgement has been made.  The order of facts that are presented represent an opportunity for bias.  Good editors and writers are able to make the best decisions on issues like these and give readers the full story.

Part of the identity of any news organization or publication is the types of stories and articles that it chooses to write.  This is a conscious decision on the part of editors and writers.  Readers learn to expect a certain type of story from different sources.  Over time, a website or publication gains a reputation and readers feel they can trust that source for certain types of information.

In the case of CNET, they have spent years developing a reputation for technology news and tech product reviews.  I often found myself reading reviews of products on CNET before making purchasing decisions or before advising friends and family on decisions.  I trusted their reputation as a knowledgeable and fair source of information.

That relationship changed last week.  The CNET editorial staff had voted the Dish Hopper DVR as the best in show product for this year's CES show, but was forced to pick something else because the executive office of CBS (parent company of CNET) decided that they didn't like the choice.  There is an ongoing legal battle between CBS and Dish over this Hopper DVR, but that doesn't mean that CBS should attempt to control the media as a result.  Additionally, CBS forced CNET to go out with a bogus official statement initially that didn't own up to the Hopper winning the vote.

I have been a very longtime reader of CNET, but this has broken the relationship and I don't see them as a trusted source any longer.