Posts Tagged: scalability

Main — Comments Off
12
Jul 10

Beliefs

Beliefs: The framework of things you hold to be true, and of which form the basis for all of your decisions.

Here are some of mine. Which do you disagree with? Why?

Workarounds are never a good thing. Short term workarounds are never short-term. They should be avoided. Do it right the first time, and if you can’t due to time or budget, delay the project. I hate technical debt.

Plan as you go is more appropriate to life and to projects, and returns better results, than planning everything up front (ie agile vs. waterfall). What we are talking about is predicting the future. Sure, you can be somewhat accurate, some of the time. But it’s just a guess. You’ll be more accurate if you don’t predict too far out. If you’re more accurate, you’ll be happier.

The problems of new are less than the problems of the old. On occasion you will run into a bug by upgrading software to the latest version. But I’ve found that on balance, I have far fewer compatibility & stability problems if I keep up to date. And as a bonus, new features!

Buy the well-built item once instead of the cheap thing multiple times. It’s eco-friendly, and you get to have the quality item to use every day. My wife and I had been wearing out a $10 garlic press once every 12 months or so with basic wear and tear — till we bought the Rösle Garlic Press for (at the time) $30. Five years later, it still looks good as new and works brilliantly.

Price is not correlated to the value. Just because it’s expensive doesn’t mean it’s worth a lot. Conversely, just because it’s cheap doesn’t mean it has no value. Open-Source Software, Wikipedia, a walk with your kids – these all have a lot of value, and they don’t cost you a dime.

Deals are rarely worth it. Everything is “on sale”. Everything is discounted. Of course, there are good deals to be had. It’s just that the effort to find and take advantage of the deal is more costly than any savings I might obtain. There is a reason why rebate forms are difficult to complete: it is in the company’s best interest that you never fill them out.

I believe in Scaling Software over Scaling People. See my blip on Techies and The Business, or the whole article here.

The most important attribute to any employee is their willingness and ability to learn. I’ve written about this one a lot. I think learning is the key to innovation, that through mistakes you get better, that Adapting is the Most Crucial Skill You’ll Ever Learn, and that progress (and who doesn’t want progress?) is an act of discovery.

So!  What are some of your beliefs?

Main — Comments Off
6
Oct 08

Server as a Service

In addition to Sun Microsystem’s MySQL, Oracle is now available through Amazon’s Web Services. One has to wonder why you would ever buy a physical database server again. This takes SaaS to a whole new level: Instead of Software as a Service, this is Server as a Service. One of the listed benefits:

Quickly and easily add computing capacity as your requirements change. Amazon EC2 reduces the time required to obtain and boot new server instances to minutes, allowing you to quickly scale capacity, both up and down, as you need.

I know it’s geeky, and I’ve probably lost most of my readers already (thanks for getting this far!), but doesn’t that sound really really nice?

Main — Comments Off
22
Jun 08

Scaling Open Source MySQL at YouTube

This is a bit technical, but most of you are in IT – A good interview over at IT Conversations:
“In mid 2006, YouTube served approximately 100 million videos in a single day. To maintain a website of that scale, one would imagine YouTube has hundreds of DBAs. But in fact, there are just three people that make it all work. Paul Tuckfield, the MySQL DBA at YouTube shares horror stories about scalability at YouTube and how he coped with them to keep the show going everyday, while learning important lessons along the way.”
I was going to outline all of the good things this says about both Open Source software, and about the MySQL software itself, but I’ll just let the interview stand on it’s own.
Main — Comments Off
26
Dec 07

Get your geek on: Scalability

Scaling software is a science – often overlooked. Everyone can fix a bug: the process is very straight forward – you replicate the bug, build a solution, prove the solution, and deploy it. But performance problems are far more nebulous. It is hard to find the problem – much less fix it once it is identified. You need performance tools, architecture expertise, full-picture knowledge of the whole system, and then the political backing to making it all happen.

But to help with my own understanding of the issues, I’ve been reading High Scalability (a site that “tries to bring together all the lore, art, science, practice, and experience of building scalable websites into one place.”) They get into some of the guts of some pretty complicated real-life examples and it is all good.

Last week they posted about 37signals Architecture (the makers of Basecamp, Backpack and Highrise.) There are some impressive numbers in there, but my favorite nuggets revealed were:

There are 2,000,000 people with Basecamp accounts.
37signlas only employs 30 servers.

If you ignore all of their other products, each server can support about 70,000 Basecamp users. That is really, really impressive. If you add in their full line or include their published goal to halve their server count in the next quarter or so, it only gets better. Don’t let any one tell you Ruby on Rails doesn’t scale. It does, and the numbers prove it.