williamhathaway.com

I’ve had my Garmin Forerunner 305 for about 6 months now and in general love it.  One thing that bugs me though is the elevation data it gathers tends to be fairly inaccurate (even in cases where the GPS signal is excellent and the measured distance covered is dead-on).  This is really easy to spot if you do an out and back course for a workout by checking your elevation data and seeing if it looks symmetrical.  If you use motionbased.com to store your workout data, you can apply their “gravity” service to have it auto-correct the elevation data on your uploaded workouts.

I did a FIRST tempo run on the treadmill tonight.  The goal was to run 3 miles @ 6:45 pace in between a warm up and cool down.  A few people had asked me what incline setting I was using when I ran on a treadmill.  I did some research and found “The 1% Incline Treadmill Myth” (and part II) which seemed to have reasonable coverage of the available studies.  After reading these I decided to not set an incline.  In a few weeks once I have more baseline data I’ll try changing the incline a bit and see what it does to my heart rate at paces compared to earlier runs.

After my warmup was completed, I started increasing the treadmill speed so I could go after the 6:45 target pace.  After flipping through the speed adjuster I saw that 8.8 MPH is a 6:48 pace and 8.9 MPH equals a 6:44 pace.  I decided I would switch between the two during the workout, trying to stay at 8.9 MPH about 3/4 of the time.  Switching the pace a tiny bit during the run helps break up some of the metronome type feel of running on the treadmill.  The first mile went by without a lot of effort with my heart rate reaching 172 BPM towards the end.  At about a mile and a half I could definitely feel heat starting to build up and I was sweating pretty heavily.  The downside of no air resistance is that you don’t get the cooling effect of the air going past you and during busy times at the gym when there are a lot of people working out on the aerobic equipment you can feel the heat and humidity start to rise.  Heart rate for the 2nd mile averaged 174 with a peak of 177.   The 3rd mile ended up with an average HR of 179.  I definitely felt like I could have kicked it in harder, but that would defeat the point of the workout.  I then dialed the speed down to a 12 minute mile and started my cool down.

I had the Garmin footpod on during the run.  I noticed that it was within 1-2% of the treadmill for measuring the distance during the tempo part of the workout.  When I switched to 12 minute and 10 minute miles during the cool down, it wasn’t reading as close (off by about 30 seconds a mile).  Viewing the footpod pace data in Garmin Training Center, I can see the data points jump around quite a bit (although that is the case with GPS data as well).

I tried running one of the FIRST speed workouts on the treadmill at the Camp Hill LA Fitness tonight.  This was a bit of an experiment since I haven’t ran anything faster than 5k pace in quite a while (months?).  The workout called for 12 x 400m @ (10k pace – 55/60 seconds per mile) with 90 seconds rest interval.  I’m currently using an estimated 10k time of 42 minutes for my FIRST calculations since my 5k times have been just under 20 minutes lately.  A 42 min 10k pace is 6:45, so subtracting 60 seconds makes the target pace for the 400m interval a 5:50 to 5:45 minutes/mile pace.

I had also recently purchased a Garmin footpod which attaches to your shoe and uses an accelerometer to gauge how fast you are running and transmits this to your Forerunner.  I really liked the idea being able to capture treadmill workout data since previously when running indoors I can only capture my heart rate and I have to remember what speed I had set on the treadmill during various parts of the workout.

I started with a warmup for .8 miles on the treadmill and the footpod seemed to be generally accurate to within 2-5% for speeds from 12 minute miles down to 7 minute miles when I was eyeballing it.  At the end of the warmup the treadmill had me listed as having covered .81 miles and the footpod told my Forerunner I had completed .80, so it was almost dead-on.  When I tried some faster speeds briefly (sub 6 minute mile) the accuracy started decreasing, but you can also calibrate a footpod to help increase the accuracy which I have not done yet.  If anyone is considering getting a footpod, check amazon.com for prices, I found it for about half of what the Garmin online store listed it at.

Now that my warmup was completed, I started a new workout on the treadmill using the interval mode.  This feature lets you program in two running speeds, and by pushing the “interval” button, it will automatically switch between them.  I programmed in 4.5 MPH for the jog speed and 10.4 MPH for the interval speed.  I jogged for a little bit and then pressed the interval button.  It took a handful (4-6?) of seconds to ramp the speed up from 4.5 MPH to 10.4 MPH.  The first few intervals weren’t too bad, but the time it took the treadmill to ramp up and down in speed as the interval was starting or ending made some of the calculations a bit fuzzy.   I was trying to figure out how many seconds in advance to hit the interval button, but after a little while I gave up on it and just accepted it wouldn’t be an extremely precise workout.  For workouts where the intervals are longer, the ramp up/down time is less of an issue.  I also realized during the workout that I should have made my jogging pace something like a 2 MPH walk in order to have better recovery during the 90 seconds rest intervals.  I started doing this after the 7th or 8th interval, which helped get my heart rate back down quicker, but the slower starting pace increased how long the ramp up time took.  I ended up calling an end to the speed workout after completing 10 intervals since my legs were feeling pretty beat even though I was still recovering aerobically.  When you run intervals outside, as you fatigue your pace naturally drops.  On the treadmill, you end up running at the set pace (or you fall off).  I’ve read that it is easier to pull a muscle during workouts on a treadmill when you are pushing it since your body’s natural mechanism to slow down isn’t available as you fatigue.

Overall I had a pretty good workout.  I think when running longer intervals the whole treadmill ramp up/down time won’t be significant and the footpod seems to work pretty well.  I checked the workout data from Garmin Training Center, and it had my max speed for intervals at anywhere from 10.4 to 10.6 MPH which is really close to what I had the treadmill set to (and certainly the treadmill may not be perfect either).  I’m going to calibrate the footpod this weekend on a track and see if the accuracy improves.

coreadm – configure the core dump facility.  You can enable/disable core dumps, control if syslog messages should occur, and what directory/file names will be used when programs dump core

dumpadm – configure how system crash dumps are handled

pargs – print out command line arguments (although only for processes ran as yourself unless you are root).  This does not have the 80 character limit that /bin/ps has.

dladm – show or change data link info (great for displaying which network interfaces you have and their status, S10)

plimit – show or change  resource limits of a running process (file descriptors, memory, run time, …)

prstat -L – shows per-thread breakdowns of processes

if_mpadm – test IPMP failover

getent - test name service routines (great for checking things out when testing LDAP name service implementations)

fsstat – show file system statistics (S10 u2 or higher)

bart – compute/compare file system checksums (sort of like a light version of Tripwire, S10)

route -p - add persistent routes (S10 u3 or higher)

One of the new features in the Solaris 10 08/07 (update 4) release is that zones can be given their own network interfaces. In previous releases of Solaris, the only networking option for zones was to give the zone IP addresses on virtual interfaces, which meant that all routing and firewalling had to be configured in the global zone. In S10 08/07, when configuring a zone, you can now specify a property “ip-type” which can be shared or exclusive. A shared ip-type is the same configuration that was available in previous Solaris 10 releases and is the default. A zone configured with an ip-type of exclusive has the following properties:

• Separate IP stack (zone admin can add multiple IP addresses, configure routing, and run snoop)
• IP Filter can be used and administered inside the zone
• All network traffic between ip-type exclusive zones goes through physical network adapters (skipping the loopback path inside the kernel)
• IPMP can be configured if there are multiple interfaces assigned to the zone

Besides networking changes, there are also a bunch of additional new features for zones in S10 08/07 (much stronger resource controls, DTrace privileges can be added, boot options, etc). Check out the latest S10 Admin Guide or the Zones Community at opensolaris.org.

I decided that it would be good to start adding some faster pace medium to long runs since I am getting closer to kicking off my FIRST training for the Lehigh Valley Half-Marathon. I’m fine with short and fast (mile and 5k races) and long and slow (runs over 14 miles), but I feel pretty intimidated by the middle ground. To help remedy this, I hit the Conewago trail late this morning figuring I’d try to get in 6 or 7 miles at around a 7:30 pace and see how it felt. I took the first mile at an easy warm-up pace in 8:54 and then started to pick it up. I hit the 2nd mile at 7:29, 3rd at 7:27, 4th at 7:24 and did a slow circle to turn around and head back. The 5th mile was a 7:28, 6th a 7:21. At this point I was still feeling pretty good so I decided to push the pace a little more. As I was getting close to ending the 7th mile I was doing some mental calcultions and it seemed like I had an outside chance of hitting 8 miles in under an hour, which I’ve only done once before (at the HARRC 10 miler, my best medium distance run yet). I ended up clocking a 7:10 7th mile and then started pushing as hard as I could and hit the 8th mile in 6:38, beating an hour by a few seconds. I felt pretty jazzed after the run and celebrated/re-fueled by chowing down on a Neato Burrito with Mark Rebuck.

Motioned Based

What: SMT Turkey Trot 5k

Where: New Cumberland Park, New Cumberland PA

When: November 22nd 2007 (Thanksgiving)

Executive Summary: finished in 19:5X, relatively flat but potentially slightly short course, race management was mediocre, post-race grub was great.

UPDATE: Official results have been posted. My clock time was 19:55 and I finished 30th out of 578 people.

Full Report: When I decided to run in the SMT Turkey Trot, I wasn’t quite sure what to expect, but I was picturing a relatively small race (maybe 50-100 people). As I was pulling into New Cumberlad Park about 30 minutes before the race was to start I realized I was very wrong. There were runners everywhere. A lot of people were jogging around warming up or waiting in line for the porta-potties and a constant stream of people were heading toward the registration site. I stood in line to register with a filled out form and my $20 entry fee. The registration process was being handled very efficiently. Volunteers had entry forms on clipboards and were handing them out to people in line who didn’t have a registration form with them. There were about 10 people in front of me when I got in line to register and it only took about 2-3 minutes to go through the line and finish the process.

I jogged around a bit and said hi to some of the Harrisburg Road Runners I recognized in the mass of people. A few minutes before the race started, I saw Nate Powell coming toward the starting line and we hung out until the race kicked off. There were a few announcements over a PA system before the race started, but it was hard to hear the announcer. I could only make out about half of what he was saying. It seemed like there was about 2 minutes worth of announcements and I could feel people on the starting line getting a bit edgy. All of a sudden a horn went off and the race was on.

The start was a bit more chaotic than usual. I suspect there were a higher percentage of people who didn’t race much as I could see a lot of people who had started out pretty quickly begin to struggle after the first 1/2 mile or so. The course had a few very moderate incline/decline sections, but overall is was pretty flat. I didn’t encounter anything I would classify as a hill, although there were a few small bumps. Traffic was being stopped by police or volunteers at a number of intersections, but there were also some places where there was no traffic control. I didn’t encounter any problems, but there were a few areas where if my timing was 10 seconds slower it could have been awkward figuring out if a car was going to wait for me or I should alter my path.

I felt fairly solid for the 1st mile (6:27), the 2nd mile I was starting to fade a bit and was wondering why I thought it was a good idea to race a 5k less than two weeks after a marathon. I was contemplated backing way off the pace and taking it in easy when I heard the two mile split being announced (13:00 according to the announcer, but in retrospect I think the 2nd mile marker may have been a tiny bit short). At that point I figured I might as well try to finish it up strong since there wasn’t a lot of distance left. At about 2 and a half miles I saw a teenager to the side wearing a full bodied turkey costume with a bunch of feathers in his hand. I heard a runner that was just behind me yell out to the turkey kid “I’ll take a feather!”. I then remembered hearing some garbled words from the announcer about door prizes and feathers, but by the time I had sorted that out in my head I was about ten seconds past the turkey kid and decided it wasn’t worth the time penalty to turn back and grab a feather.

My stomach was really hurting for the last half a mile or so, but you could see the finish from 400 or 500 meters away and having the end in sight makes it a lot easier to deal with the pain. I’m not sure of my exact time, I think the official clock was reading in the low 19:50s. My watch time was 19:49, but I waited until I crossed the starting line before beginning the timer, so I was probably a few seconds short of the official time. Garmin had clocked my 3rd mile at 6:15, and a additional .08 miles in 25 seconds. I suspect the race might be a tiny bit short, but there were some sharp corners during the race where the Garmin might have lost a few feet due to sampling gaps. Either way, my pace was right around a 20:00 5k and I’m fine with that.

After walking down the long chute (a good idea for the large number of entrants in the race, apparently there were some problems last year with people cramming up in the chute) I handed the tear off section of the bib to a volunteer and wandered over to the food area. This race had an awesome spread of food. There were bagels, pastries, pie slices (blueberry and pumpkin), cookies, coffee, hot chocolate, water, soup, breads.

I met Nate after he finished (his hip had been hurting him so he ran a very conservative race) and we hung around a bit eating some of the goodies and trying to figure out where we could see the results. Unfortunately, even with waiting around for about 20 minutes we couldn’t find the typical results board that you see at most races, or even a person we could ask about the official places/times. I hope the results get posted to the web site soon.

Race kudos: The race-day registration process was handled very well and all the food laid out was fantastic.

Race improvements: It would be nice if the results could be posted on a viewable board or printed out soon after people finished.

Motion Based

I received an email from RoadID.com about a new “interactive” option and went to check it out.   The interactive option is essentially a web site that lets you key in a bunch of additional information (medical history, health insurance, extra contacts, etc).  If you choose create a Road ID (RID) with the new method, on the RID, there are a few lines for the standard information and then there is contact information for the website and an 800 number that can be called in the event a first responder needs the information.  On the back of the Road ID is a serial number and pin that need to be provided by the first responder to access the information.  This lets you provide quick access to a lot more data (blood type, allergies, etc) and is relatively secure (well, as secure as their website is).    The only thing that was a bit disappointing to me is they charge $9.99 a year to  maintain the data after the first year.   Given the very low volume of data and extremely infrequent access  I would think they could offer  the data maintenance quite a bit cheaper.

I joined some friends and coworkers in signing up for the Lehigh Valley Half-Marathon (LVHM) which will be held in Allentown, PA on Sunday April 27th.  I ran the LVHM that was held in April of this year, it was the first “big race” of my adult running career and I had a fantastic time.  The race was very well put together, I would highly recommend it to anyone looking for a half-marathon this spring.

I’m going to train for the race starting in mid-December using an 18 week half-marathon program from the FIRST people.  Hopefully this will let me build some strength on top of the mile/5k speed I’ve developed.  I bought and read the book and have a few thoughts on the program from analyzing the paces and workouts suggested, but I’ll wait until I’ve been training on the program for five or six weeks before commenting.

If you are using scp to move files between machines where one or both sides have a slow processor, you might find significantly increased speeds if you specify that the cipher used should be blowfish instead of the default ‘3des‘. (This is for cases where single-threaded CPU hits a bottleneck before network or IO limits become a constraint.) When copying large files from my MacBook Pro to a 1ghz Niagara box I see the throughput increase from ~ 4.5-4.7MB/sec to ~ 7.1 MB/sec.

Use the -’c’ option to specify the cipher.

Example: scp -c blowfish file.tar $user@$host:/path/to/file.tar