How to get a whole bunch of jobs going while everyone else is sipping eggnog
While everyone enjoys a break from work this time of year, one thing that won’t be taking a break is grid computing. GPGrid will run at full capacity at all times, as will many of the usual offsite computing clusters. We encourage users to continue to submit jobs so that they can run over the holidays.
Support will also be at its normal levels on all days that are not Fermilab Holidays. Enjoy the holidays and, as always, take advantage of all available computing resources. With universities going on semester breaks soon, now is a good time to test out your offsite submissions if you haven’t done so yet, as we may see more opportunistic resources than usual in the next few weeks.
Some guidelines for job submission:
1. Submit jobs at a rate not exceeding 1K jobs/minute (so wait at least 10 minutes after submitting a cluster of 10K)
2. Submit no more jobs at a time than could reasonably run within a few days to a week. If your experiment allocation is 1000 slots and your cluster has 5000 8-hour jobs that use less than 2 GB of memory, were you to get the full allocation of your experiment, then you should expect the cluster to conclude in approximately 40 hours. But you will rarely get the full allocation, and so 80 hours would be a more accurate estimation.