Demonstrating we can run on the bulletq
Using GR 20-08-11 and running the default Gleam unit test to produce the resulting merit file using the RHEL5-64 boxes and then again on the bulletq (RHEL6-64). Attached is the log from both the RHEL5-64 and RHEL6-64 runs as well as the jobOptions.txt. We then ran Luca Baldini's meritDiff.py on the resulting merit files. There were some differences - most are Aud*** variables - so are associated with the running times, which are unsurprisingly different. Other differences are noted here: