Disqus: Scaling the World’s Largest Django Application

http://ontwik.com/python/disqus-scaling-the-world%e2%80%99s-largest-django-application/

by ahmicro • 15 years ago

101 31 15 years ago

ontwik.com

grovulent4• 15 years ago

Not like this is a problem I have to worry about. But where on earth does one learn this stuff?

The talk is useful - as an overview of what they use - but I know nothing of how to implement a single step.

sushi• 15 years ago

I think most of it comes from experience, sticking to fundamentals and learning from the mistakes of others but don't listen to me. I'm a noob to all this stuff myself.

buro93• 15 years ago

It's called experience.

Which perhaps sounds rude, but it's not meant to be.

This stuff isn't taught per se, you learn it bit by bit as you solve each problem that you face.

I learned about HAProxy when my site load exceeded that which a single web server could manage.

I learned about heartbeat when I had to update my HAProxy and it knocked the site offline.

I learned about master/slave replication of databases when a site I worked on had considerably more reads than writes and scaling vertically (buying a bigger box) cost more than scaling horizontally (adding cheap read slaves).

I learned of sharding when I worked on a graph stored in an Oracle database and performing calculations on the whole graph exceeded that one physical box.

I learned of one-hop replication to solve the problem of a sharded graph.

I learned of partitioning to solve the problem of having one big database and the computability not being maxed but the storage being maxed.

I learned of memcached when I wanted to reduce page generation times and realised going to the database was more expensive than keeping it in cheap RAM elsewhere on the network: http://www.buro9.com/blog/2010/11/18/numbers-every-developer...

I learned of reverse proxy caches when I wanted to make sure requests for things already served never reached the web layer again.

I learned about Varnish when I considered that most reverse proxies use disk storage for their cache.

We can go on and on here, but the message is that you learn these things one at a time solving real problems that you come up against. There is always the next hurdle to jump through, and when you get there you too will learn how to get past it.

I'd emphasise that you cannot attempt to do this prematurely, that premature optimisation quote really applies well to architecture too. Keep things as simple as they can be and just know that when you get to a hurdle that someone else has already solved it and you've just got to find out where it's written down (if anywhere), what they used, how they approached it, the upsides, downsides, what they'd do differently, etc.

You could try sites like http://highscalability.com/ but I would urge you not to implement things without knowing why you're implementing them. Don't cargo cult ( http://en.wikipedia.org/wiki/Cargo_cult ) this stuff, it's really key to do only what you need to do, when you need to do it.

grovulent4• 15 years ago

seriously dude... write that book...

Cause i can't find it anywhere...

mmt• 15 years ago

This stuff isn't taught per se, you learn it bit by bit as you solve each problem that you face.

Some of it is taught. I learned about HAproxy second-hand, before any problem had to be solved.

I do agree that much of this has to be learned initially through experience, but I also believe much of it can be taught. Teaching it in a classroom wouldn't work, though, as Ops only exists when there's something to operate.

It's my perpetual dilemma: few startups have enough [1] scale to warrant having a senior enough sysadmin to be a mentor, let alone another one or two more junior ones, but large companies are awkward environments for experimentation.

I try and dump some bits of wisdom/experience to my blog. Shameless plug: http://blog.maxkalashnikov.com

[1] Even Disqus is borderline in terms of traffic. Moore's law and even its more linear corollary for I/O can take one a long way.

fields• 15 years ago

I find in practice that even more important than figuring out which tools to use to solve your problem is figuring out what your problem is in the first place. The importance of measuring tools is often vastly under-represented in this discussion. Doing so properly almost always involves writing some custom profiling code, and having a good understanding of where resource bottlenecks are likely to be in your systems (and then ignoring that when the numbers tell you otherwise).

patio112• 15 years ago

I have some experience with optimizing, scaling, and administering less-than-web-scale systems. Much of the knowledge comes from reading posts like this or presentations (thank God for Slideshare), or attending conferences where they give talks with titles like How We Scaled Linkedin (that was JavaOne if I remember correctly, and that talk was single-handedly worth the price of sending me to America for my company).

You can also learn from observation when working with people who are better than you. This is one of the best reasons to have some industry experience prior to raring off into startup land. I am a much, much, much better engineer in 2011 than I was in 2007 partly because I am older and wiser but mostly because I sat next to the second best engineer I've ever met and took notes when he told me that everything I knew was wrong.

Then there is the last option: your lack of expertise with X bites you in the keister, you fix X, you now have expertise with X and hopefully write up your experience somewhere to decrease the net amount of keister-biting in the world.

andrewcamel• 15 years ago

Thank you so much for submitting that. I'm creating a Django application that has the potential to store and work with even more data than Disqus, so I've always been worrying about how to scale this to such a huge scale. Thanks to your submission, I'm no longer as crazed about it.

yuvadam2• 15 years ago

Disqus is a classic case study when it comes to scalability.

If there is one thing I learned from Disqus it is the power of keeping a lightweight stack. Disqus keep it simple, and prove that all the myths that "Django/SQL/whatever doesn't scale" are obscene.

Even for an app with requests per second in the 5 digit range - they do pretty damn good with the basic Django stack with no more than some small tweaks.

St-Clock• 15 years ago

That was exactly my thought.

No NoSQL. And they use transactions for write queries.

They use apache (not even nginx!).

Only 25% of their servers are pure (no snapshot) caching servers (not 50% or 75%).

They prefer vertical partitioning over sharding (but they still use sharding).

I understand they are looking at redis for some of their features, but really, their main stack is traditional and proven. Very enlightening.

ghc2• 15 years ago

I actually find it astonishing that they're using Apache for this. I have had a lot of problems with Apache behaving weirdly in the past, especially with mod_wsgi.

dansingerman• 15 years ago

I know disqus does loads of traffic, but sheesh ~100 servers - thats exceptionally non-trivial.

moe• 15 years ago

On the upside, there's long phases of "just add more of the same" in scaling these things. The fun comes in waves, every time when you hit one of the various physical barriers (namely latency and bandwidth).

That is to say the host-count in isolation is not the most interesting figure. I've seen large sites run on 20 machines or on 500, depending on the skills of the management- and developer-team, and how much they care about the infrastructure cost in the big picture.

The host-count becomes more interesting when you relate it to the request rate. 17k/sec is absolutely a worthwhile workload, even when (as likely in the disqus case) reads dominate writes by far.

That said the relation of 100 hosts / 17k rps seems about reasonable.

However (not meaning to narrow their achievement) the engineer in me can't help but wonder if perhaps even a little more could be squeezed out on the caching front? I was a bit surprised to not see varnish on the slides; fragment caching on the perimeter can achieve mind-boggling results.

thedz• 15 years ago

Varnish wasn't in production at that point. We're testing/using Varnish now for some things. It definitely is helping.

In the general caching front though, I want to note though that Disqus is particularly hard to cache -- there's a very long tail leading to relatively a high miss:hit ratio per pound of caching.

sugarcode• 15 years ago

Great points. We've actually deployed additional caching on the frontend, and Varnish should be appearing on future slides.

nas• 15 years ago

As an aside, AFAIK douban.com is still using Python and Quixote[1]. Back in 2007 they were doing 2 million pageviews per day[2,3]. According to Alexa they are busier yet now. They use the SCGI protocol as well.

1. http://quixote.ca/

2. http://mail.mems-exchange.org/durusmail/quixote-users/5441/

3. http://mail.mems-exchange.org/durusmail/quixote-users/5657/

zeemonkee• 15 years ago

Good to hear Quixote is still going. It was my favourite framework in the pre-WSGI Django/Pylons days.

adrianwaj• 15 years ago

How does Disqus make money? Has IntenseDebate made money since being acquired? Does revealing this information make Disqus a more likely acquisition?

ro_gupta• 15 years ago

Disqus has premium add-ons: http://disqus.com/addons

Disqus: Scaling the World’s Largest Django Application

Related Stories