Bump mocha timeout to 10000 to mitigate potential jenkins issues

[benbp]

Proposing a potential way to address what could be jenkins timeout issues when loading modules in mocha "before" blocks.

@RackHD/corecommitters @tldavies @davidzuhn

[brianparry]

@benbp One of the issues here is that each test calls startServer in a before block and endServer in an after. We need to investigate why, but swagger takes several seconds to initialize which means that tests require long timeouts. I think this is a reasonable workaround, so 👍 but we should still investigate if there is a way to reduce swagger startup time or reduce its impact on test time.

[tldavies]

👎 This should try to be fixed via jenkins with lowering the node count for the build slave first. As it is not breaking on RackHD. So I think this is the build machines fault. More comments on RackHD/on-core#128

[benbp]

@brianparry yeah I was wondering if perhaps there was some way to cache whatever loading swagger does so we could avoid the test lag.

Closing this for now since @tldavies is suggesting the root problem with exceeding timeouts (rather than just taking a while) could be in the provisioning of the build environment.