[TRAVIS] Increase timeout for health checks

[sehqlr]

• Installs cleanly on a fresh build of most recent master branch
• Upgrades cleanly from the most recent release
• Updates documentation relevant to the changes

We originally added a two minute timeout to the health checks to allow
the cluster time to become healthy. Lately, there are a couple of checks
that may need more than two minutes, so we are going to increase it to
five.

[ryane]

lgtm. the do build failed due to a timeout. it doesn't look like this change contributed to that since the build never even got to that point. it looks like a package install was taking too long or hung.