Update the stress factor for CIs

[k-sareen]

Given that the meaning of stress factor has changed in #198, the CIs need to be updated to reflect that. Currently as it stands, all CI jobs time out as they use MMTK_STRESS_FACTOR=1. I'd suggest we increase the value of MMTK_STRESS_FACTOR from 1 to something like 1000 or 10000 (maybe even more depending on how long it actually takes) ASAP.

[qinsoon]

We do want to keep stress factor as 1. Currently the stress test timed out for JikesRVM (3h for each benchmark/test), and the stress test failed in around 1min for OpenJDK.

We should see how long it would actually take to run with STRESS_FACTOR=1 for JikesRVM. I think it may be able to run some tests with STRESS_FACTOR=1.

For OpenJDK, we need to fix the bugs.