Though this is the second day of SAF, it is really the first official day.  Yesterday was actually a pre-conference day.

Today, we had three keynote speakers.  Neil Leslie and Tim O’Brien from Microsoft, and Bob Pearson from Dell.  One interesting thing about the keynote speakers at this event is the that they were all limited to 30 minutes.  When hearing this, I initially felt that they would be too time constrained to be able to deliver a complete message, but all three were successful.

After the keynotes, the follow-on break out sessions, and lunch, we all gathered together for an Open Space session led by my friend, Steven “Doc” List.  This was my favorite part of the day (excluding some one-off conversations that recurred throughout).

Open Space, in a nutshell, has gets everyone together to nominate topics for discussion.  Once the topics are decided and organized (I am REALLY skimming over this part), everyone attends the topic discussions that they desire to hear or participate in.  I don’t know how many formats Open Space “supports” during topic deliberations, but I experienced two.  One topic I attended, the use of social networking in the enterprise and the challenges it presents such as privacy and security, was setup as chairs in a circle for discussion.  This was a good and productive session, but it was nothing compared to the first one I attended.

The first session I attended, “I don’t get Twitter”, had too many people to put all the chairs into one big circle.  So it was arranged with a “fishbowl” in the center.  The fishbowl consists of four chairs are placed in the center of the room in a circle facing inward.  There are some very basic rules here.  No one is allowed to participate in the discussion unless they are sitting in the fishbowl.  Only three people are allowed to sit in the fishbowl, meaning that there must always be one empty chair.  As someone wants to enter the discussion, they enter from the audience and sit in the empty chair.  One of the original three, usually the one that has been there the longest, must exit the fishbowl.  Once out of the fishbowl, you return to being part of the audience and can no longer participate in the discussion without reentering the fishbowl.

The “I don’t get Twitter” session was packed!  There were 10 topics run in parallel for a conference of about 300 attendees and there were over 50 people in this session alone.  There were many interesting questions posed during the session.  “What about security/safety of our children as they expose personal and geographic information about themselves?”, “How is Twitter monetized?”, “Does Twitter detract or enhance the work environment and should employers care?”, “What happens when a company is built on Twitter and Twitter ‘goes away’?”. 

They were all very good topics of discussion and there were many participants.  I personally, was addicted to the fishbowl and found myself reentering it several times during the session.  As you may have expected, the Tweeps in the audience were tweeting about being in an Open Space discussion about Twitter during the discussion.

The session lasted 50 minutes and was exhilarating.  I have asked Doc to facilitate our Open Space at the Atlanta ITARC in February and now that I have participated in Open Space, I am extremely excited.

I was the last person in the fishbowl (go figure) and I announced, as everyone left, that Doc had just tweeted to me “have you explained Twitter yet?”

To follow me or Doc on Twitter our aliases are jdecarlo and athought respectively.

And there you have it.

Joe.

The trip to San Francisco was uneventful, but the arrival was beautiful.  I was tipped off that I should get a window seat to witness the landing of the plane into the San Francisco International Airport.  It was quite beautiful.

My first day was very fun.  In addition to the planned festivities, I got to spend time with quite a few people that I normally don’t get to see face-to-face.

Of all the first day “Pre-conference” presentations, my favorite was by Ulrich “Uli” Homann.  Uli is a Principle/Partner Architect at Microsoft and is a Infrastructure/Solution Architect combination which is a rare breed.

Uli spoke to us about the energy crisis in data centers and how application design and virtualization can work together to address the issue. 

When analyzing how the industry addresses Green computing, the trend is for more efficient CPUs and lower power rack densities.  While cooling is a big energy consumer in today’s data centers, disk utilization of mass storage does not get the attention it deserves and consumes more energy than CPU, memory, and fans combined.

To win in this space, there are two basic approaches that can be taken.  Stabilizing resource utilization, while increasing service throughput, where services represent any server output.  And stabilizing that service throughput, while decreasing resource utilization.  While both these can address the problem, Uli presented solutions leveraging the first approach.  

An example of this would be planning around a stable storage constraint.  Let’s take a mass storage solution that houses 100 terabytes and takes 25 servers to support servicing that storage to consumers.  Stabilizing service throughput while reducing resource utilization would be accomplished by increasing each server’s efficiency so that the same 100 terabytes can be serviced by 10 servers with the same throughput.

Uli warned against over execution of cautious planning.  That is to say, plan for High Availability (HA) or Disaster Recovery (DR), but don’t implement it if you don’t need it.  Having huge amounts of unneeded replication in our data centers is contributing to the energy crisis.  Since the government is putting restrictions around the amount of power all data centers can consume (a percentage of the country’s overall power consumption), the days of paying for servers or storage that we don’t need in the name of “just-in-case” need to come to an end.  As architects, we need to be more responsible in identifying what deployment configurations we actually need before requesting them.

Uli introduced us to a concept called “Constraint Based Planning” which he picked up from a book by Eliyahu Goldratt called The Goal: A Process of Ongoing Improvement.  I am not going to try to define it here, but he illustrates it using the data center and the comparison of energy spent versus the service units as you can see from the diagram on the right.

He suggests, from a infrastructure and solutions perspective, that services and applications should be segmented so that administrators can segment the work loads of servers.  All that is to say that if a service isn’t needed, don’t deploy it.  If it has to be deployed but is rarely used, don’t waste energy making it immediately available, but create a topology that will allow it to be serviced up from a hibernated or serialized and stored state.

This idea can be translated to application development, which is depends on, to allow application features to be turned “off”.  This is to say that not all features are immediately available.  Today, applications typically expect service calls to immediately respond.  To help solve this problem, applications should have resilience to failed service calls and allow for retry logic when it is not available.  So if the data center has “turned off” a particular service by some sort of hibernation, but can bring it back online when a request for it occurs, the application can support this kind of delay.

Though this session was mostly attended by infrastructure architects, there was a lesson for solution architects as well.

This does very little justice to Uli’s presentation, but, hopefully, there is something for you to digest here.

Comments welcome and encouraged.

Joe.

This past Wednesday, I attended a general bike maintenance class at REI.  This is a free class and I would recommend it to any beginning rider.

To kick off the class, we covered changing a tire.  The instructor, Cat, demonstrated on a road bike tire, but assured us that the principles she was going to show us applied to several different types of tires including mountain bike and hybrid tires.  Though stressing that a better practice is to bring a spare, she proceeded to show us how to patch a hole in a tube.  This exercise was thorough and complete.  Here are some of the tips she shared with us:

• Identify where the hole in the damaged tube is, even if you are going to replace it, so that you locate the area on the tire for inspection.
• Ensure that the tube has a little air in it before trying to install it.  This will keep the tube from twisting and make it easier to manage.  A tube that is installed with a twist, once inflated, can rupture or have other problems.
• If possible, refit the bike tire using only your hands, as using tire levers to perform this task risks damage to the tube.
• Before fully inflating the tire, inspect the entire tire to insure that the tube will not get pinched between the tire and the rim.  This could either cause the tire to lose its bead, or the tube to get a pinch flat.
• Fill tire to proper pressure.  When on a trail, there are two basic ways of inflating your tire.  A hand pump or a CO2 cartridge.  It is tempting to put “just enough” pressure with a hand pump.  Be sure not to under inflate the tire.  Most tires have minimum and maximum tire pressures listed on the side wall.  Make sure you achieve a tire pressure that falls within that range.  A lot of tires max out at 120psi and CO2 cartridges dump about that pressure into the tire when fully engaged.  Be cognizant of this and check your pressure before you ride. 

Next we covered derailleurs.  A quick definition of derailleurs.  Most bicycles have multiple “gears” and the mechanism that changes these gears is the derailleur system.  The derailleur system consists of a front and rear derailleur.  Both consisting of a chain and several sprockets.  The front derailleur has a guide called a cage that is used to move the chain between sprockets derailing it.  The rear derailleur has what’s called a jockey pulley to control the derailing of the chain between the rear sprockets.  Both the front and rear derailleurs have adjustment screws that control the high and low gear limit stops.  The limit stop is the distance that the cage or jockey pulley allow the chain to go horizontally along the sprocket grouping.  A maladjusted limit stop can either allow to chain to go too far, causing it to fall off of the sprocket group, or not far enough, causing it to not shift into the highest (or lowest) gear.  When adjusting derailleurs, you should also consider the amount of tension the cables have.  Too loose or taught cables can also affect the shifting of gears.

I am by no means an expert in this area, so I will not attempt to explain how to adjust derailleurs.  In fact, some REI store locations offer a Derailleur Adjustment class.  In Atlanta, the Perimeter location is the only one that offers it ($30 for members/$50 for non-members).

Finally we covered basic maintenance.  How to care for your chain and other components.  Here are some of the talking points:

• All Allen Bolts on bicycles are metric.  It is very easy to mistakenly use a standard Allen Wrench on them, thus stripping them.  Only use metric Allen Wrenches.
• Keeping the components that have moving parts properly cleaned will extend the life of the component.
• Use a degreaser to clean the chain.  She recommended  Finish Line Speed Degreaser.  Ensure that you do not ride the bike after degreasing without first lubing the chain.
• Allow the degreaser to completely dry before applying lube to the chain.  She recommended Tri-Flow Superior Lubricant.
• When lubricating the bicycle, find all parts where metal touches metal and add lubricant to that area.  A single drop of lubricant is usually enough.
• Do not forget to add lubricant to the cables where they meet the housing.  These cables are called Bowden Cables and the plastic sheathed housing is lined with metal where it contacts the cable.  Be sure to place the cable on the smallest sprocket when applying the lubricant as the smallest sprocket causes the least amount of resistance on the cable, thus putting it in the most exposed position.  Then change the gears to the largest sprocket (progressing to the easiest gear for the rear derailleur and the hardest gear for the front derailleur).  This will increase the tension of the cable causing it to get pulled through the housing dragging the newly applied lubricant with it.

One quick note that was covered in the class was to be cognizant of chain wear.  The chain is the quickest thing to wear out on a bike and is often overlooked until it breaks.   She recommended the Park CC-2 Chain Checker though REI doesn’t carry it.

And there you have it.

Joe.

Yesterday, I cashed in on my wife’s Birthday present to me: a Bike Fitting at Atlanta Cycling, Vinings.

Before I dive into the details, let me explain what a bike fitting is and why they are a good idea.  Many people purchase bikes.  Road bikes, mountain bikes, time trial bikes, hybrids… all kinds.  The people who purchase them from mediocre bike shops should expect that the sales person will ask you your height and possibly the length of your pant inseam, if they ask you anything at all.  From this they will determine what size bike you should purchase.  Some places will even have you stand over the bike to determine the clearance between the top tube and your body.  Sounds good, right?  I thought so, too.  Fortunately, most good bike shops have competent employees that know at least the minimum measurements needed to purchase the correct size bike for a person.  But that is just getting the right sized bike…that doesn’t fit the bike.  A bike fitting is a series of measurements and observations that tell the bike fitter how to adjust both the bike components and you, the rider, to get a better, more comfortable fit on the bike.

The reason I felt I needed to get fitted was because my arms would fatigue far faster than my legs and my breathing.  Occasionally, my hands and my left foot would fall asleep during rides.  I stumbled upon someone else getting fitted at Atlanta Cycling and I watched for about 10 minutes.  I was sold.  I learned so much just watching someone else’s fitting that I knew I would greatly benefit from getting one of my own.

As instructed, I brought my bike, to the shop dressed in full riding gear.  After mounting my bike onto the bike trainer, Mike, the fitter, observed as I rode for about five minutes.  He asked me to make a few small adjustments to the way I was riding (i.e. hand placement) while he observed.  After he was satisfied that he saw enough, he explained that he was going to record me during some interval riding.  He started the recording and I rode for about 3 minutes.  He had captured my video into a program on his laptop.  Now he could do analysis of my riding posture in slow motion and using the program’s measurement utilities.  Once he established a baseline, he started to the first of the many changes.

Starting “from the ground, up”, he wanted to adjust the cleats on my shoes first.  He has me walk in a straight line, and the return to toe a line he made with masking tape.  Measuring off the natural position of my feet while standing with more tape, he used this information to adjust my cleats.  After he was done, I rode the bike again.  Using a tripod-mounted laser pointer, he pointed them at my knees.  We could both see that my knees now went smoothly up and down during each rotation.  Before the cleat adjustment, my knees made tiny loops at the top of the rotation.  This sounds like a small change, but I could tell a big difference.

Next he decided to work on getting my heel lower.  Apparently, I, like many people, “toe pedal”.  Toe pedaling is when you point your toes too much while pedaling.  After asking why this is bad he replied with an analogy.  Imagine that you were sitting on a balance beam with nothing to support you but a set of 2×4’s just close enough to touch with your toes.  This would be fine enough to take the pressure off of your perineal nerve until your toes and calves got tired.  Eventually, you would rotate back off of your toes and put most of your weight on your perineal nerve.  Not only does this cause discomfort, but could put a damper on future fatherhood, if one so desired.  Keeping the heel parallel to the ground for as much of the rotation as possible engages larger muscle groups in the leg and thus is more power and less fatigue.  He fixed the problem, mostly, with seat adjustment.  He raised my seat a total of about 2 inches.  Interestingly enough, the bike shop that sold me my bike said that I may have to lower my seat some more, but it would involve cutting the seat post since it was as low as it would go.  Initially, I could tell the difference of the seat adjustment, but it put more pressure on my arms.

Next he addressed the added pressure to my arms.  First he changed the stem of my handlebars.  It raised the handlebars significantly.  The pressure of on my arms disappeared, but I still wasn’t fully comfortable.  He measured my shoulder width and then my handlebars.  I didn’t realize there is a correlation, but my shoulders measured 44 inches and my handlebars measured 40 inches.  I wasn’t sitting natural on the bike because I was reaching inward two inches on each side of my body just to grab the handlebars.  Replacing the handlebars and the stem turned my Bianchi into a totally different bike. 

After a few more adjustments and measurements and a lot more observations, we were done.  I rode around the parking lot and couldn’t believe the difference that all the little adjustments made.  He then recorded me doing intervals again.  He overlaid the first, baseline, recording over the new recording.  You could see many differences.  One was my head placement.  My head was higher and it was more comfortable for me to look forward, so I did it more.  My heel position was lower to the ground and the angles were less aggressive.  I was able to sit up higher in the saddle.  One huge difference I saw was the raising and dropping of my hip during rotations.  Initially, they were so dramatic that you could see my biking being pulled from side to side as I strode.  Now it was minimized and the bike was stationary.

This was a great experience for me, and I would recommend it to anyone who is serious about riding.  I am sure that there are a lot of places that do this, but if you are in the Atlanta area, visit Atlanta Cycling’s Vinings store.  Ask for Mike, he’ll do a good job.

And there you have it.

Joe.

      I am sitting here watching Superbad for like  the 20th time and I decided to comprise my top ten favorite stupid movies.

Before I get to my list, I think it is important that I clarify what I mean by stupid movie.  I don’t like movies that are just stupid and are based only on the merits of being stupid.  I like movies that are borderline slapstick, but not in a Mel Brooks kind of way.  One essential criteria a good stupid movie must have is that it must be quotable.  The plot can be half-retarded (because you never go full retard), the actors can be complete no-names, and the script can be atrocious as long as it has good quotable lines.

And now for my list:

10	9	8
7	6	5
4	3	2
1

And there you have it.

Joe.