What to know for the September 2017 GPGrid Refactor¶
- Monday, September 18, 5 p.m. CT: jobsub client will deny submissions starting
- Tuesday, September 19, 8 a.m. CT:
- Any running jobs will be killed
- All grid jobs submitted with jobsub client that have not completed by this time will be killed, whether running at Fermilab or offsite
- Tuesday, September 19, 9 a.m. CT dCache will become unavailable (Tevatron Run II, Public and CMS dCache)
- Wednesday, September 20, 3 p.m. CT: Planned end of outage for GPGrid. Note that you can submit jobs, but dCache will NOT work until the end of that outage.
- Wednesday, September 20, 8 p.m. CT: Planned end of outage for dCache
- The GPGrid cluster will be renamed to FermiGrid. If you specify sites to submit to in jobsub_submit commands, you will now have to use FermiGrid instead of FNAL_GPGrid.
- you would specify the site like this now:
jobsub_submit --site=FermiGrid -N 1 <additional options>