Upgrading issues with seadas

[OB.DAAC - SeanBailey]

Hmm...
Following your steps, I was able to replicate the problem.  A short term fix is to move the getanc.py call after the modis_GEO.py call - you don't need to run it until you want to do L2 processing..

We'll look into why the att/eph search is now unhappy if done after the ancillary data search.
FYI, modis_GEO.py calls modis_atteph.py, which grabs the appropriate att/eph files and both modis_atteph.py and getanc.py use the same local sqlite database.

You can get processing going for files already grabbed by call in modis_atteph.py with --refreshDB, e.g.:
$ modis_atteph.py -v A2019240165500.L1A_LAC --refreshDB
Sean

[woodbri]

woodbri@u19589217:/u$ ./install_ocssw.py --version
install_ocssw.py 3.0
woodbri@u19589217:/u$ uname -a
Linux u19589217 4.4.0-151-generic #178-Ubuntu SMP Tue Jun 11 08:30:22 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux

[woodbri]

FYI, I was using "getanc.py --refreshDB A2019240165000.L0_LAC.L1A" in my script. This might have been doing something to break GEO loading data via modis_atteph.py.

I modified my process-seadas script so it runs "modis_atteph.py -v file.L1A --refreshDB"  before modis_GEO.py and then runs "getanc.py --refreashDB file.LIA" after that and that seems to have gotten by the reported problem.

Thank you for the assistance and awesome support you guys provide.
-Steve

[gnwiii]

Colleagues have been using scripts that download product files and associated ancillary files and process to qkm L1B.  When all the L1B files are, done the level-2 processing jobs run in parallel.  Internet access is poor, so it is not unusual to have some downloads fail, which could change the order in which modis_atteph.py and getanc.py are run and might explain intermittent failures by modis_GEO.py.  If so, the issue has existed for months, maybe a couple years.

[OB.DAAC - SeanBailey]

I believe I have found (and squashed) the bug.  'Twas in anc_utils.py - which is used by both getanc.py and modis_atteph.py.  There was a file check on the local sqlite database that didn't include the anctype in it's call.  It now does and seems to have done the trick :grin:

Update and let me know if anything goes awry.  (yes, I did test it :wink:)

Sean