Date
01
Goals
- Begin imaging with GPUs
Discussion items
| Item | Who | Notes | |||
|---|---|---|---|---|---|
| VLA Lustre upgrade | James | https://staff.nrao.edu/wiki/bin/view/NM/VLALustreReplacement | |||
| GIbson/NMT network | James | Peter is adding a dual port card to gibson. 129492 I think he reconnected the fibers wrong. gibson cant get off campus. | |||
| OOM error for some torque jobs | James | Twice now VLASS imaging jobs have been killed by torque for going over memory. This shouldn't ever happen. | |||
| HERA orders | James | Ticket to rack herastore01d 127013 Ticket to rack herastore02 and shelves 129289 | |||
| herastore01 | krowe | Done: five missing/bad disks on VD_3. helpdesk ticket 127465
Done: Mar. 4, 2021: rebuild herastore01-4, tell Paul and he will put data back on it and let Librarian know about it. Remove backup copies krowe made of the shelf (herastore01a) | |||
| nmpost091 | krowe, Jrobnett | Done: nmpost091 is up at NMT (127446 and 126684). Should we change the hourly rsync to be every other hour so it doesn't conflict with itself as often? Nevermind. It is seeing jame's rsync for his job and exiting. Perhaps I should just remove the check or make it smarter. | |||
| Return VLASS nodes to torque | krowe | Done: Jobs are draining off so we can start returning some nodes. 127669 Keep one node like nmpost071 but not in condor and offline. | |||
| Tobin imaging step | James | Need to add imaging step to htcondor workflow (works, waiting for word to roll into git repo) pbmask=0.4 Should krowe fold step25 into the git repo? | |||
Lustre oss 127011 | |||||
| Investigate VIP vs pipleline results | James | VIP script and pipeline with w=1 are generating images that vary at around 1e-5 | nmpost004 | krowe | jtobin is borrowing nmpost004 on Jan. 21, 2021 |
| Jira | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|
|
...