Chutney often fails because Tor hasn't bootstrapped yet

Travis appears to be under heavy load right now, so we're getting about one bootstrap failure per pull request: https://travis-ci.org/torproject/chutney/jobs/517006783 https://travis-ci.org/torproject/chutney/jobs/516995533

There's two ways to fix this issue:

1. allow 120 seconds for bootstrap
2. make the network run with 10 second consensus intervals

Option 1 is easy. Option 2 might lead to some instability, but it's worth trying.