Here’s a little Christmas video I put together with some help of some friends.

I’ve always had a huge problem when it came to backing up my personal data – it costs too much to do it right. Every time I buy some “system” for backing up, it’s always woefully inadequate. I’ve purchased tape drives, optical drives, and external hard drives in the past and it’s not possible to keep enough backup media around to ensure you can recover from a bad backup with the rate at which hard drive sizes are growing. Unfortunately, I’ve had an urgent feeling (call it geek’s intuition) the past few months that I really need to backup my data somehow. Enter Mozy, an online backup service for Windows and Mac (beta) acquired by EMC in October 2007.

I found this list of “Alternatives to Mozy” on their website. It’s funny because I’ve done most of them.

• Burn a new CD or DVD every Sunday night and store it at your brother-in-law’s office.
• Pay $200/year for an online backup service that uses old, mediocre software.
• Buy a $200 external hard drive and hope your office doesn’t burn down.
• Do nothing and don’t worry about backup. (We suggest closing your eyes, plugging your ears and repeating “I’m in my happy place, I’m in my happy place.”)
• Run a cron job of rsync, gzip and mcrypt piped over ssh to your friend’s server over his DSL line.

So how does it work? Using the Mozy software, you define what data you would like to backup by directory, file type, last modified, etc. It even comes with a very comprehensive list of backup sets already defined. Once you setup a regular backup schedule, Mozy takes care of the rest. In the background, it will encrypt your data with 448-bit Blowfish encryption and send it to the Mozy servers via 128-bit SSL. You can either use the default Mozy key or supply your own, but if you lose your key your data is lost forever since no one will be able to decrypt it. If you use their key, they will not view the data you backup or sell your information (see privacy policy).

Once Mozy does the initial backup, it works in the background (set it and forget it) to detect files which have changed and sends only the blocks which have changed over the network. This will save storage space on their end and time and bandwidth for you. The good news is that both the CPU time used for encrypting and the network bandwidth can be throttled easily and flexibly. The Mozy software can even backup files which are currently open (locked) and backs up Outlook files (.pst files). Essentially, the whole thing is a block-level continuous incremental backup.

I initially signed up for the free MozyHome service with a 2GB quota just to try it out. It’s perfect if you just had a few files that you wanted to keep safe like your most important documents. However, I quickly found out that this would not be enough since my wife has 26GB just in digital photos she’s taken. The initial selection of files from the default backup sets came up to 89GB (I was able to trim this down to 69GB with some additional filters). Luckily, there is a MozyHome plan which is $4.95/mo per computer (you get 1 month free if you pay a year in advance and 3 months free if you pay two years in advance) for unlimited storage space. Yes, I said unlimited.

So Mozy’s been running in the background for a while and disaster strikes. How do you recover your data? If you’ve just accidentally modified a file you didn’t intend to, you simply right-click on the file in Windows Explorer and select “Restore Previous Version” and you can restore any version within the last 30 days. You can also access the “MozyHome Remote Backup” virtual drive in windows explorer and browse the files on the Mozy servers before restoring. If you need to restore files to a computer which doesn’t have the Mozy software installed, you can use the web-based restore. Finally, if the ultimate disaster strikes and you need a copy of all of your files and don’t want to download them, you can request that your data be burned to DVDs and shipped to you overnight. There is a fee for this service, obviously.

I already have an increased peace of mind knowing that my most critical documents and folders were backed up when I was trying out the 2GB quota. Now that I’ve gone to the unlimited, I will feel even better knowing that all of our family photos, music collection, and home videos will be safely stored with a company that is an expert in data storage.

Blogged with Flock

Tags: backup, online, emc, mozy, mozyhome, privacy, unlimited

Things are winding down here at CMG and I took most of Thursday to work on an assignment that is due this weekend. It’s been a fun week despite being sick, but I’m ready to get back home.

Millicomputing: The Coolest Computers and the Flashiest Storage

This talk was presented by Adrian Cockcroft and discussed the idea of using very low-power technology such as that found in cell phones to build a cluster of lots of servers in a very small space. Each individual device use less than a watt (milliwatts) so hundreds of these things can be packed into a very tight space.

The other advantage is that these device would use flash storage which has very good transfer speeds with no seek time and no rotational latency. The only disadvantage is the number of writes are finite so these devices would have to be used for most read-only applications.

In terms of cost, the average device can be purchased for less than $100 and a 6GB high-capacity high-speed SD card only costs about $75. For the cost of a typical full-sized server you can purchased hundreds of these things.

Consolidated Capacity and Performance Reporting

This is my session which is starting in about 30 minutes. Wish me luck!

Then it’s off to the airport….

Blogged with Flock

Tags: cmg, cmg2007, millicomputing

Things are winding down here at CMG and took most of Thursday to work on an assignment that is due this weekend. It’s been a fun week despite being sick, but I’m ready to get back home.

Millicomputing: The Coolest Computers and the Flashiest Storage

This talk was presented by Adrian Cockcroft and discussed the idea of using very low-power technology such as that found in cell phones to build a cluster of lots of servers in a very small space. Each individual device use less than a watt (milliwatts) so hundreds of these things can be packed into a very tight space.

The other advantage is that these device would use flash storage which has very good transfer speeds with no seek time and no rotational latency. The only disadvantage is the number of writes are finite so these devices would have to be used for most read-only applications.

In terms of cost, the average device can be purchased for less than $100 and a 6GB high-capacity high-speed SD card only costs about $75. For the cost of a typical full-sized server you can purchased hundreds of these things.

Consolidated Capacity and Performance Reporting

This is my session which is starting in about 30 minutes. Wish me luck!

Then it’s off to the airport….

Blogged with Flock

Tags: cmg, cmg2007, millicomputing

Here’s a summary of the sessions I attended:

Application Performance and How Apdex Makes It Better

This session was about application performance management as a component of Business Service management (Infrastructure Service Management being the other component) and how the Apdex spec can help. The most interesting aspect of Apdex is how simple it is. If you can measure some component from the perspective of the end-user and determine at what point the end user is satisfied, tolerating, and frustrated then you can calculate an index of application peformance. The index goes from 0 to 1 so it is fairly easy to compare results from many applications and determine which ones might need some attention.

In order to measure an application using apdex, you need to determine the point at which end users “feel” the impact of a performance problem. If the application is an average web-based application (e.g. a blog), then most users feel a website is responding well if the page returns in under 2 second. Between 2 seconds and 8 seconds, most users will continue to use the site but may flip to other things while waiting. Beyond 8 seconds, most users will give up and go on to something else.

Of course, this will vary by application. One theory put forth is that the time a user will spend on a given page will determine how quickly they get frustrated. If the page is a login page or some information with a “next” button, then they expect a quicker response time than if it’s a blog post that they will spend several minutes reading.

The formula is: Apdex = Satisfied Count + (Tolerating count / 2) / Total Samples. What it boils down to is that you get “full credit” for satisfied responses, “half credit” for tolerating, and no credit for frustrated. If taken as a percentage, the three values always add up to one (see the Apdex Cases session for a neat use for this fact).

Once you have an Apdex score, you need to determine what is a “good” score and a “bad” score. Goals can be set to improve the Apdex score as part of the Application Performance Management process. Typically, a score of .94 to 1.0 is consider Excellent, .85 to .93 is considered Good, .70 to .84 is Fair, .50 to .69 is Poor, and anything less than .50 is unacceptable.

Apdex Process

This session covered more information on Apdex and how to develop a process for incorporating it into an overall APM (Application Performance Management) strategy. Keep in mind that traditional capacity management (Infrastructure) of the servers, network, and storage is still important. APM simply gives you another angle of attack to the problem. Each user “flows” through the infrastructure in a different way and just looking at individual components of the infrastructure is not enough.

Some of the different kinds of vendor tools which can gather Apdex metrics and/or generate Apdex reports. This includes active monitoring, passive monitoring, and end-user agents. Also discussed was potential pitfalls to using these types of monitoring when new technologies like WAN accelerators are used.

The Best at Better

The Plenary speaker was Dale Dauten, author of several nationally syndicated newspaper columns and several books. He was not the kind of speaker you would expect at a technical conference, but I think this is what made the talk so good. He asked everyone to think to think of someone who they consider to be the best boss they ever worked for and think of what makes that person the best. Through his storytelling, he gave us some really good things that make those kind of people who they are.

The Laughing Warriors: How to Enjoy Killing the Status Quo

This breakout session was also hosted by Dale Dauten. It was in a much smaller room and he was better able to have a more intimate discussion about the concepts he covered in the larger Plenary session. After some “warmup” exercises, Dale asked us to break into groups and think of our favorite place to visit and why. Could we run our businesses or departments like that place? For example, our group picked Six Flags because it’s a fun place to go even when you have to stand in line and you always leave feeling like you didn’t do everything there was to do. Indeed, you want to go back.

Setting Performance Objectives Using Apdex

This session was all about how to determine what “T” should be for an application. This is a critical factor for determine the Apdex score. Peter Sevcik gave us 10 different methods to choose from as possibilities. They ranged from going to the users desk and using a stopwatch while watching their body language for signs of tolerating or frustration to calculating a mathematical inflection point (don’t ask, I didn’t understand this one).

Apdex: Case Studies

Three different case studies were presented of companies implementing Apdex reporting. The first case study was from AOL and covered how they initially began using Apdex to monitor various web properties and the refinements and tweaks they made to the process over time. The second was a case study from Keynote discussing using Apdex with a VoIP provider to monitor call quality with metrics such as call clarity and delay. The final case study was the use of the Barycentric Coordinates system to display lots of scores for applications in a visually appealing and compact way. This topic was also covered in the Barry session I attended earlier in the week. Good stuff.

Blogged with Flock

Tags: cmg, cmg2007, apdex, dale dauten, apm

Things are really moving along now. Lots of sessions to attend, not much time between them. Here is my summary of the sessions I attended today:

Statistics for Performance Analysis & Capacity Planning

This session was actually presented in 3 separate sessions each an hour long by Ray Wicks from IBM. I’m probably at a bit of a disadvantage to many of my colleagues because I have never taken a statistics class or had to work much with them. It was very enlightening to get an introduction to many of the concepts in statistics, at least the ones that pertain to Performance Analysis and Capacity Planning. There was still a lot of the Greek alphabet, but I think I walked out of the session with new and valuable information.

To Instrument or Not to Instrument, That is the Question

This session covered instrumenting code (regardless of language) to make application performance management much more effective. The focus was on “always on” metrics that could be useful for keeping tabs on what’s happening. I think the main point was that by the time you know you have a problem, it’s too late to change the code to add instrumentation and increase the probability of introducing new bugs. The happy medium is somewhere between no logging and leaving debug turned on in production.

Death to Dashboards

This was probably one of the more highly anticipated sessions of the conference so far and has certainly been the most entertaining. The main theme was that displaying information on “traffic light” dashboards can only indicate so much information. For example, that server shows red but “how red” is it? How long has it been red? How important is the system? The presenters, Peg McMahon and Justin Martin from Sprint Nextel presented some of their experiences in implementing a capacity and performance monitoring system. The alternatives presented went along with some of the other data visualization techniques discussed in other sessions. The one I had not seen previous was Treemaps – a neat way of visually displaying lots of information in a small space using different sizes and colors. It reminded me of tag clouds in a way, but more graphical.

ITIL v3 Capacity Management: A Review

I’m not sure what I expected to get out of this session. I’ve heard ITIL mentioned numerous times in the past few years, but I don’t think anyone really gave me a one or two sentence summary of what it’s for and why you would use it. I guess I have a little better understanding now, but a lot of the references in this session were assuming you already knew a little of ITIL v2. Still confused a little, need to do some more research on the concept.

Coming on Friday

Before I close this summary, I thought I would post the abstract for the session I’m presenting on Friday:

Since I now have a full-blown cold, I woke up early and unable to breathe which makes it hard to sleep. So, I figured it was a good time to take some sunrise photos from my 15th floor hotel room overlooking the Marina. You can’t see them, though, because I sent them into Cowpie Buzz for her photo contest. I’m sure they’ll show up there for you vote on eventually.

Since the conference didn’t officially kick off until the afternoon and I was up early, I decided to have breakfast in my room. I just have one question, who thought asparagus for breakfast (or anytime) was a good idea?

After breakfast I spent some time catching up with work and actually made some pretty good progress on my paper which is due this coming Sunday. It wasn’t a lot of progress, but certainly more than I’ve made to date – combined. I hope the block I’ve had getting started on it is broken now.

For lunch, I headed back down to Seaport Village to eat at a place called Buster’s Beach House and Longboard Bar. I opted for the Pork Luau sandwich which was really good, but it took longer than I would have wanted for the food to be delivered. Then again, I’m impatient like that.

Keynote

The keynote was given by Jerred Ruble, President and CEO of TeamQuest Corporation and was entitled “Is Capacity Planning Still Relevant?” Obviously, the answer at a conference for capacity planners has to be “yes”, but he had some very interesting things to say about how capacity planners (read: IT geeks) must learn how to frame capacity and performance metrics in terms of business metrics. As tools and techniques have evolved, it’s more critical than ever to convince executive teams that capacity planning is still relevant.

After the keynote there was a panel of former A.A. Michelson Award winners discussing the next steps beyond Best Practices which were dubbed “Next Practices”. I thought it was very interesting and telling that when asked how many in the audience had Myspace or Facebook pages, hardly anyone raised their hands. When the same question was asked about the audience’s children the story was very different.

I’ve already made the observation that the CMG conference still seems to be attended by a lot of Mainframe types and could do with some attention from the social networking and Web 2.0 crowds. The keynote did shed some light on this, but the approach seemed to be for the normal attendees to embrace the new technologies rather than go out and recruit the new generation of developers and bring them into CMG.

The expectation is that the hottest topic for next year will be “Green IT”.

Using Simulation to Forecast Performance: A Case Study

This was an interesting extension of some of the modeling and Queuing Theory covered in yesterday’s workshop on PDQ. This presentation seemed to be lacking in substance and referred a lot to the paper itself, but I understood more than others I think because I attended the PDQ workshop. To the author’s credit, she was presenting a case study and not trying to cover Queuing Theory.

Seeing It All at Once with Barry

This session was the most interesting of the day. It was presented by Dr. Neil J. Gunther and Mario Jauvin, both of whom presented in the two workshops I attended yesterday. The concept was about Barycentric coordinates and other ways to create ways to present large amounts of data in a visual way. This is similar the stuff being done at Gapminder.org and Digg Labs‘s Swarm.

Business Meeting and A.A. Michelson Award

I usually don’t bother attending the Business Meeting, but I knew the buses were leaving for the harbor cruise right after and I didn’t want them to leave without me. I’m glad I attended, it was definitely a positive experience to see the inner workings of CMG and the work that goes into putting on the conference. It makes me think about volunteer besides just writing a paper and presenting for future conferences.

Congratulations need to go out to all of the various award winners this year, but especially Adrian Cockcroft for winning this year’s A.A. Michelson award. His contributions to CMG and the field of performance management, particularly for Solaris are well-known to most in the industry. I think the most interesting part of his acceptance speech was his advice in presenting metrics to Executive Management: metrics should be presented as dimensionless or in terms of dollars or they will try to convert the to dollars. Sometimes they even try to convert dimensionless metrics to dollars as well. I also enjoyed what he called the “Cockcroft Law of Book Writing“.

Adrian also touched on a talk he is giving later in the week on something he has dubbed “millicomputing” because the power consumption must be measured in Milliwatts. These are very small, low power devices with flash storage (limited write capability, but very fast reads). However, Adrian’s idea is to use arrays of these small devices to create a very powerful, low power cluster. Unfortunately, I think I’m going to miss the talk but I’ll have to check out paper once I get back home.

PARS (Performance Analyst Relaxation Session)

This was a don’t-miss event. As far as I know and according to everyone I’ve asked, this is the first year they have had a outing for PARS. Usually PARS is food, drinks, and entertainment in the convention center. This year they moved the food, drinks, and entertainment to a ship, the Inspiration Hornblower, as it cruised around San Diego Harbor.

I would have gotten more pictures, but the battery in my phone was low. We cruised out beyond the large bridge you can see in the photos taken from my hotel room, turned around, and cruised back. We even passed a large battleship as it was heading back out to sea. Speaking of which, everyone was talking about touring the USS Midway and now I have to make time to go see it.

 

Blogged with Flock

Tags: sandiego, cmg, cmg2007, inspiration hornblower, buster’s, seaport village, barry, barycentric

The CMG Sunday workshops are an opportunity to get a more immersive and hands-on experience with Performance and Capacity Planning topics. Since it all happens before the official kickoff of the conference on Monday afternoon, it’s also a great way to get the jump on the other attendees.

In the morning, I attended a workshop by Adrian Cockcroft and Mario Jauvin called “Performance Management with Free and Bundled Tools”. They have actually presented this same workshop for several years now and its updated each year with the latest information on dozens of tools. It was this workshop in 2005 that led me to implement several open source tools and write my own paper about the experience which I’m presenting at this year’s CMG.

For the afternoon session I opted to sit in on Dr. Neil J. Gunther’s “How to Move Beyond Monitoring, Pretty Damn Quick!”. This was a very enlightening session for me because it discussed how to move from merely monitoring performance to modeling and predicting performance. Previously, I’ve looked at using statistical means of predicting performance such and there are obviously some drawbacks and limitations to this. Namely that statistics assume that response time will grow in a linear fashion. However, a computer system is really a circuit of queues and Queuing Theory tells us that it will be more like a hockey stick. PDQ is a an open source tool which allows you to build queuing models and use them to predict performance. For example, if my server farm has an average response time of 300ms with a load of 1,000 requests per second then what will performance look like at 6,000 requests per second? At what point does response time become unacceptable and I should add additional capacity? I’m looking forward to exploring this further once I get back home.

After the workshops, I set out to find someplace to dine that wasn’t terribly expensive and since I’m limited without a rental car and don’t want to take taxi, I decided to explore Seaport Village some more. At the far end, I found a place call the San Diego Burger Company. The price was more reasonable than other places in the area, but still expensive for a burger joint. However, it was definitely worth the price. The burger was great, but the onion rings as far from crispy as you can get.

On my way back to the hotel, I came across a unique site. A carousel inside of a building.

I realized as I was walking back to the hotel that because of the time zone change, my class assignments which are due at midnight EST would now be due two hours earlier here on the west coast than at home. I hope my professors will understand why I was an hour late turning in my assignment. I’m looking forward to a bit of relaxation and a chance to work on my paper since there’s nothing scheduled until Monday afternoon. It seems that Monday evening’s social/networking event is a harbor cruise which sounds fun.

Tags: sandiego, cmg2007, cmg, seaport village, performance, pdq, free, open source, san diego burger company

After a bit of stress this morning with getting to the shuttle on time and getting my prescriptions refilled at the last minute, I have finally made it to San Diego.

I won’t even cover the security process, it’s not worth it. Let’s just say that I don’t fly often enough to remember the rules and procedures. The TSA folks are so kind to let you when you screw up, though. Trust me.

The shuttle driver gave me a gate, the self-check in kiosk confirmed the same gate, but then i got a flight status notification from American Airlines (set them up last night) saying the gate was different. So, thinking that must be the most up-to-date information, I headed down there and watched the plane above load and take off for Ontario, CA. I was early, so I figured my flight was just the next one at the gate. Then the gate agents closed up shop and left. The waiting area emptied and the monitors at the gate desk went blank. Curious, I got up and went to the bank of monitors that listed all of the flights and the gates. Wow, I was at the wrong gate.

So, I get down to the correct gate and it hasn’t started boarding yet. Finally we get on board and leave the gate about 15 minutes later than the scheduled departure. Other than that, it was really an uneventful flight. We hit a little bit of turbulence, but it wasn’t too bad.

If the Palm trees don’t give it away, this is from inside the airport terminal looking out in San Diego. I had a little trouble finding where to catch a taxi, but the ride to the hotel was uneventful. I don’t usually do things like valet parking and use the bellhops at hotels that have them, but for some reason I was ready for someone else to carry my luggage today.

My room is on the 15th floor of the hotel. Look to the left and see downtown San Diego and to the right is the harbor (not sure of the name of it yet). Nice view, huh? I’ll try to get some night pictures and post them as well. Once I got up into the room I discovered that I lost the stylus to my Treo somewhere along the way, probably when i was carrying my bags on my shoulder and they kept trying to knock the phone out of the belt clip. However, the clock radio by the bed has an iPod dock built-in. How cool is that?

Around 8:00pm, I headed down to the registration to get my packet with my badge, swag (they call them collectibles), meal tickets, final agenda, and book for the Sunday workshops. I scoped out the restaurants in the hotel itself and determined that they were either too formal (I was wearing jeans and a t-shirt) or too expensive or both. I got a list of restaurants within walking distance and found a sign to Seaport Village. There was a restaurant called Harbor House. A little expensive, but I had the fried shrimp, fried sea scallops, and fried fish platter with lobster bisque soup. Delicious.

Now it’s time to get to bed early so I can get up in plenty of time to get breakfast and find my way around. Just a bit more reading to do for my class assignments since I have a few things due tomorrow. Oh, and I can’t forget that monster of a paper that’s due next weekend. I MUST find time to make significant progress on it this week.. like even finish it.

Blogged with Flock

Packing for San Diego trip

The time has come for my trip to San Diego for the CMG conference. I will also be presenting my paper on Friday at 11 o’clock. I’ve also decided to use this trip as an opportunity to experiment with moblogging and live blogging. I’ll be posting some here, some to twitter, and some directly to my Facebook page. I even signed up for a service called Utterz which allows you to call in and record your voice then TXT video, photos, and text and it will combine them to one entry and post to a blog, twitter and facebook profile.

More tomorrow, gotta finish packing and getting caught up on Honey-Do’s!