John reminded us that PIMScache and calibration need a lot of disk space so have to be run at DSOC.
Timeline for SECI aw-project workflows - running at RADIAL nodes?
Disk space issues in Socorro will drive how much outside compute we can use.
DAs currently working on automation project to try to speed QL, target of 75% images passed without looking at them. Speeding up QA will help reduce latency on disk.
Also need to look at disk space needs more carefully, e.g. testing creates 80 SE cont imaging sets. Likely that upcoming 100TB extra will help, as will finishing SE cals, but not clear it's enough.