terraref / computing-pipeline

Pipeline to Extract Plant Phenotypes from Reference Data
BSD 3-Clause "New" or "Revised" License
21 stars 13 forks source link

restore clowder instance for terraref #221

Closed robkooper closed 7 years ago

robkooper commented 7 years ago

The clowder instance for terraref had a catastrophic disk failure.

max-zilla commented 7 years ago

Import is still in progress. The flirIrCamera is currently uploading and has been since Saturday afternoon, but this is the most complex/largest collection to recreate - this is one of the longest file lists, and each flirIr dataset only has a single file in it so there are more datasets created than, e.g. ps2Top which has a long file list but 100 files per dataset so it processes much more quickly.

RAW DATA

EVALUATE

LEVEL 1 DATA

dlebauer commented 7 years ago
dlebauer commented 7 years ago

@robkooper can any data be recovered from the database? For example, the UAV / Tractor data? Or the sample data that Max had prepared?

ghost commented 7 years ago

Repo IDs have changed Sample data in the database may be able to be restored

max-zilla commented 7 years ago

@robkooper we should look at:

...to see if we can recover them before trying to rebuild.

@craig-willis @dlebauer are eager to see if this is possible.

max-zilla commented 7 years ago

Space IDs from my local backup of Terra production:

max-zilla commented 7 years ago

flirIrCamera finally finished at 3AM, and we determined from memory patterns we should be safe to run 2 threads on this at once. cropCircle and ndviSensor running now.

max-zilla commented 7 years ago

cropCircle and ndviSensor still running, but I updated extractors to be more robust in recognizing when their outputs already exist on Roger. This meant I could start re-deploying extractors without waiting for the raw_data rebuild to be totally done. I've started doing so.

code updated; just need to run on VM

more work to do:

max-zilla commented 7 years ago

note: make sure old LAS files have been merged according to new extractor logic from #237

ghost commented 7 years ago

@max-zilla please update.

max-zilla commented 7 years ago

I've been updating my first comment as this has progressed, although at some points I prioritized ingesting new data so others could access before getting back to this rebuild.

priSensor is about 1/3 done, then we are 99% done. the last few sensors have a small amount of data to restore comparatively, and we might restore parts of the Level_1 data but others have been subsequently deprecated by updated extractors or workflows (e.g the plot level stuff)