Connection with boto to AWS hangs when running with crond

I have a very basic python script which uses boto to query the state of my EC2 instances. When I run it from console, it works fine and I'm happy. The problem is when I want to add some automation and run the script via crond. I notices that the script hangs and waits indefinitely for the connection. I saw that boto has this problem and that some people suggested to add some timeout value to boto config file. I couldn't understand how and where, I added manually /etc/boto.cfg file with the suggested timeout value (5) but it didn't help. With strace you can see that this configuration file is never being accessed. Any suggestions how to resolve this issue?