June 2009
6 posts
CouchDB naked
Few months ago I spent sometime browsing through the entire CouchDB source code because this seemed like the best way to learn how large-sized Erlang (and specifically OTP) apps get written. Other choices were to look at ejabberd or RabbitMQ but they were a bit intimidating because you’ve the added complexity of understand the XMPP and AMQP protocols. I’ll be doing a series of blog...
“In the application, at mps-expenses.guardian.co.uk, each MP’s expenses and...”
– The breakneck race to build an application to crowdsource MPs’ expenses | Politics | guardian.co.uk
Amazing what can be achieved at a newspaper company by hiring a small bunch of smart web developers. So far I’ve only heard of NYT in the USA and Guardian in the UK doing this. Just wish...
Running live queries against Hadoop's offline data... →
LinkedIn folks show how they built a simple on-disk format/storage-plugin for their KV-store, Project Voldemort in order to do fast lookups on Hadoop’s output data. Huge amounts of output data from Hadoop jobs are converted into this special format via another hadoop job. This data is then dropped straight onto the Voldemort nodes and key-lookups are performed instantly.
I was thinking of...
From JavaOne 2009: Load-Testing Clustered... →
Interesting discussion on load testing by Terracotta’s co-founder, Ari Zilka.
“Processes sending messages to a process with a long message queue are penalized...”
– from O’Reilly’s “Erlang Programming”
decorators and doctests
Just ran into this issue while adding some tests to the Redis Python client. If you’re using doctests, and the method being tested is wrapped in a decorator, then your doctest for that method won’t be visible to the doctest runner. This is because when you wrap a func/method in a decorator, it loses its docstring unless you set it back explicitly. In Python 2.5, you could simply avoid...