Archive by Author

There will be a Nexus Two

17 Jul

Even if it’s not called “Nexus.”

Android, the Open Source project, is free for all to use. It’s not just limited to those companies in the Open Handset Alliance. Unfortunately, this leaves Google with the problem that they cannot always control what the carriers do to the platform. Their only means of control is forcing them to compete with hardware made “with Google.”

The crapware is invading Android. Carriers are busy trying to prevent owners controlling their own hardware. HTC adds Sense. AT&T has so crippled the Aria, the only way to install custom applications appears to be though Windows-only software. AT&T even brazenly swapped Google search for Yahoo.

Users don’t care about a lot of this, but it drives developers nuts. And Google knows that where the developers go, users follow. Google is anxious to get as many innovative applications written for Android as it can, which it needs developers to do. So Google has been releasing its own hardware ever since the G1.

Google will likely continue to do just that, keeping good, open hardware in the hands of their developers. If you’ve ever wondered why developers of the cyanogen mod aren’t frequently hired into the Google mothership, it’s because they don’t want to destroy the community. A vibrant developer community is central to making Android the dominate mobile platform, and thereby cementing Google as the mobile search of choice.

Lastly, while Google will certainly continue making their own developer phones, it’s unlikely to be still called the Nexus because they were denied the trademark.

Twitter change and why Leemba is awesome

14 Jul

Ok, so my last post on Leemba and Twitter wasn’t entirely complete.

Everything is showing a major change, even the number of alerts generated:

The original image shows a big drop in response time. But if you look closely at the options there, you’ll see I set the “sample method” to minimum. This is the aggregate Leemba is using to trim the number of points. Twitter is being checked every minute and the time period shown is about 10 days, so that would be around 14400 points to graph.

If Leemba displayed all of those points, then a chart like this would look like a big block of color. It’d be impossible to see a trend in all of that noise.

Here’s the original:

Most monitors will simply display averages for this kind of historical data, but Leemba offers a number of tools to help analyze charts and data. Here’s the same period, but with the average sample method. This is generally what you see in other monitoring solutions.

This looks roughly the same (although the scale has changed). It doesn’t tell us too much we didn’t already know.

This chart has another option, however. By checking the “Show Standard Deviation” checkbox, we can get a sense of the variability of the graph. Here’s the same data again, this time with the standard deviation displayed (right axis).

With the standard deviation displayed, we can see we’ve got an awful lot of variation, even increased variation when the minimum response times have gotten faster.

Interestingly, the standard deviation (stddev) never really drops below a second before the change, but after there are periods with much less. What this might have been caused by was the opening of a data center closer to the monitor, which might lower response times but and error rates. Or it could have been a change in the geographic load balancing.

The variance (sometimes a request is fast, sometimes it’s slow) could only be really explained by somebody very familiar with the Twitter architecture, which is why you need tools like Leemba that can display data in different ways. If we wanted, Leemba can zoom into any of these time periods on the fly, which makes it fantastic for testing performance changes in real time or evaluating the impact of past changes.

Twitter upgrades?

9 Jul

It certainly looks like something has changed.

See the image below from Leemba:

Interesting.

Update: Leemba on Terracotta

26 Jun

Just a quick update to this post: it’s pretty awesome.

I’ve spent some considerable time adjusting Leemba for Terracotta, mostly running down missing locks and various performance tweaks. With about 20 services per second, I started at a load average of over 25. But now the load is negligible. :-)

Of course, the culprit was calling out to the native fping. I was a bit surprised, I figured Linux would cache fping and it wouldn’t be quite so terrible. So now there’s now a preference to use Java’s isReachable() instead.

I’m happy with performance for now. It’s down to no more than 20% cpu utilization between two clients and two Terracotta servers. That could be vastly improved by working to use multiple schedulers, but for now I’m waiting to see what the next release will do for performance.

Project Management should go to Boot Camp

1 Jun

My time in the Marine Corps taught me a lot about leadership. They did that by teaching me to tie my shoes, to swab the deck and how to make the tightest hospital folds. Project management could learn from boot camp, too.

My first night I had to stand firewatch (guard) at 0200. I woke, stood my post and as I was being relieved I was ordered to go “swab the deck.” The Drill Instructors didn’t tell me what a swab was or where to find the soap. I was to figure it out and to not bother them about it.

That’s a common pattern in the Corps. There’s the mission and, by God, figure out a way to accomplish it. Contrast that to many civilian companies, where it’s normal to “program in Word,” which means to write the program in the requirements. The documents not only describe the result, but every painstaking step to reach that goal. Or companies suffer the impossibly long Gantt charts with fictional duties and hourly estimates. They describe the how and not the why.

Barney

It reminds me of something the Corps calls “Barney style.” Whenever we failed, we’d be reminded just how “freakin’ retarded” we all were, and then they would walk us through every procedure step by step. That’s how a recruit learns to tie their shoes. One lace at a time, left over right. Barney style.

Sadly, too many companies only ever use Barney and never leave the implementation details to the programmer. That way the chief architect gets their way, but nobody ever progresses past infantile stages. Nothing new is ever attempted and no chances taken. While it does guarantee a certain base line, all the shoes do get tied, but it also suppresses even better solutions.

Barney style should only be used in the most trying cases. Instead, the Corps has many ways of identifying what it calls the “end state.” This is where you should end up after a mission, what’s supposed to happen. Most of the time, as long as the end state is accomplished, then little else mattered. If you had to get creative, then you were simply being a good Marine.

We had all kinds of phrases that meant a Marine got creative, my favorite being ”crack deal.” That’s a couple Marines who, unknown to their superiors, worked out some deal to get the job done. Contrary to the name, it was often a good outcome. As Stanley Kubrick noted, the Corps doesn’t want robots. They want a professional, aggressive Marine who’ll use all of his abilities to learn and adapt.

Companies should want that, too. Otherwise companies resign themselves to the knowledge and expertise of just a few. No matter how well intentioned or brilliant, they can never cover everything. Instead, overly detailed requirements guarantee a suboptimal solution — every time.

Worse, usually only the most important projects warrant a project manager. In fact, the more visible a project is, the more chaffing the requirements become. That’s completely backwards.

Senior programmers shouldn’t need detailed instructions. The more junior the developer, the more direction and supervision they may need. Whether or not the project is important doesn’t have any effect on a developer’s ability to follow through on a project.

Update: ESPN gets it? “Upon talking to those involved with the projects discussed here, we got the impression that the bigwigs basically let the technologists run their own show (within reason, of course).”

I heart Froyo

27 May

image

I had to get a picture of this last night. The wife and I were at Froyo Life, not to be confused with the awesome Android release.

Dropping the IE6 project

24 May

I got quite a bit of interest on the article, a handful of emails and zero real leads. In most cases, we seem to be talking about very large corporations where you could never just setup access — even to a development or QA environment.

Oh, well. I have a ton of other things to do. It looks like those companies that are still using IE6 are content (doomed?) to keep using it.

Leemba running on Terracotta

22 May

Over at leemba.com I’ve added some posts that I’m using sorta as running notes on how to setup Leemba on Terracotta.

I’m also deploying with two nodes next week at work, and shortly after that I’ll be adding 5,000 ping services. That might not go super well considering java doesn’t (really) support ICMP. I am currently calling out to fping. Meh, I guess we’ll find out.

Have an IE6 only application?

17 May

IE6 needs to die and I’ve decided to do something about it. The trouble is, I need applications to test and most legacy apps will be only on a company’s intranet. And since I’ve already applied this method to my company’s apps, I need volunteers.

If you have an IE6 application that I can test with, then please drop me a line at mike@migrateie6.com. Please give me a short description of it (ActiveX, broken Javascript or browser detection, etc) and how I’d test (VPN or public URL).

I offer in return the easiest route to migrating your company to IE8 and later Windows versions.

**Update**

Ouch. I had to clone and move the Leemba demo off my main slice. It was pretty starved for memory and my load average went over 10. It took me 20 minutes or so to clone and reconfigure everything–slicehost rocks!

And yet, I’ve only gotten one email so far. The places I’ve posted so far cater to developers and startups, who probably aren’t locked into IE6. So if you know somebody, please send them my way. Moving off IE6 will be good for everybody. :-)

Android tablet running Air/Flash

5 May

Want.