Naama Kraus wrote:
Hi,
I'd like to know whether there is any particular recommendation for an IDE
when developing Hadoop. I did manage to checkout Hadoop core trunk and
compile it using ant.
How do folks work ?
Thanks, Naama
1. There's no specific IDE mandated. your choice of eclipse, netbeans,
idea or emacs. I use Idea and love it, except that my laptop needs much
more memory.
2. There is a bias towards unix for running hadoop nodes; if you want to
debug bits of hadoop -or code in a job- you really need to be able to
run hadoop in the local machine, hence a unix/linux platform is good.
3. I use Apache Ivy (under ant.apache.org) to manage sharing hadoop
artifacts with other projects on my system. There's a patch in bugzilla
to add this alongside the normal build files.
4. I haven't integrated it into our CI system. We are transiting off
Cruise Control and onto Hudson-on-Virtual Machines. Sometimes tests time
out on Hudson that dont fail anywhere else, which I think is clock
related -if the VM doesnt get enough CPU time, the tests time out.
-steve
