Closed sverhoeven closed 3 years ago
Full disk is likely culpred
df -h /data/volume_2/
Filesystem Size Used Avail Use% Mounted on
/dev/sdb1 50G 50G 20K 100% /data/volume_2
The global 5min pcrglob parameter set does not fit the 50Gb cache disk
du -hs /data/volume_2/vfs/dcache/parameter-sets/pcrglobwb_global/*
49G /data/volume_2/vfs/dcache/parameter-sets/pcrglobwb_global/global_05min
544M /data/volume_2/vfs/dcache/parameter-sets/pcrglobwb_global/global_30min
24K /data/volume_2/vfs/dcache/parameter-sets/pcrglobwb_global/rhine_05min_era5.ini
and real set is
du -hs /mnt/data/parameter-sets/pcrglobwb_global/global_05min/
231G /mnt/data/parameter-sets/pcrglobwb_global/global_05min/
I set the rclone vfs-cache-max-size
from 500Gb to 48Gb and stopped rclone mount, cleared cache, started rclone mount.
@Peter9192 should I create a new VM with a 500Gb cache disk?
Restarted notebook and got past 2002-07-29. Cache dir became full and then some file got deleted and it filled again
root@localhost:~# df -h /data/volume_2/
Filesystem Size Used Avail Use% Mounted on
/dev/sdb1 50G 50G 908K 100% /data/volume_2
root@localhost:~# df -h /data/volume_2/
Filesystem Size Used Avail Use% Mounted on
/dev/sdb1 50G 49G 1.9G 97% /data/volume_2
root@localhost:~# df -h /data/volume_2/
Filesystem Size Used Avail Use% Mounted on
/dev/sdb1 50G 49G 1.8G 97% /data/volume_2
A fix for running nb is in https://github.com/eWaterCycle/infra/pull/96
Notebook runs but pixel is incorrect as hydrograph shows flat line around 0, added custom model location in https://github.com/eWaterCycle/TerriaMap/blob/ewatercycle-v8/wwwroot/init/ewatercycle.json
Notebook now runs ok with nse of -0.61
Failed at simulated date 2002-07-20 with
"Exception calling application: Input/output error"
Log of run
``` PCR-GLOBWB (PCRaster Global Water Balance) Global Hydrological Model Copyright (C) 2016, Ludovicus P. H. (Rens) van Beek, Edwin H. Sutanudjaja, Yoshihide Wada, Joyce H. C. Bosmans, Niels Drost, Inge E. M. de Graaf, Kor de Jong, Patricia Lopez Lopez, Stefanie Pessenteiner, Oliver Schmitz, Menno W. Straatsma, Niko Wanders, Dominik Wisser, and Marc F. P. Bierkens, Faculty of Geosciences, Utrecht University, Utrecht, The Netherlands This program comes with ABSOLUTELY NO WARRANTY This is free software, and you are welcome to redistribute it under certain conditions See the LICENSE file for more details 2021-09-03 06:46:14,888 configuration INFO Model run started at 2021-09-03 06:46:14.877912 2021-09-03 06:46:14,888 configuration INFO Logging output to /home/sverhoeven/6b45df02-3c59-4123-b6bc-80d951729ff5/ewatercycle_output/pcrglobwb_20210903_064604/log/pcrglobwb_ewatercycle.ini_2021-09-03T06.46.14.877912.log 2021-09-03 06:46:14,888 configuration INFO Debugging output to /home/sverhoeven/6b45df02-3c59-4123-b6bc-80d951729ff5/ewatercycle_output/pcrglobwb_20210903_064604/log/pcrglobwb_ewatercycle.ini_2021-09-03T06.46.14.877912.dbg 2021-09-03 06:46:14,888 configuration WARNING The option "useMODFLOW" is not defined in the "groundwaterOptions" of the configuration file. We assume "False" for this option. 2021-09-03 06:46:14,888 configuration WARNING The option "maximumDailyGroundwaterAbstraction" is not defined in the "groundwaterOptions" of the configuration file. This run assumes "0.050 m/day" for this option. 2021-09-03 06:46:14,888 configuration WARNING The option "maximumDailyFossilGroundwaterAbstraction" is not defined in the "groundwaterOptions" of the configuration file. This run assumes "0.020 m/day" for this option. 2021-09-03 06:46:15,151 currTimeStep INFO number of time steps: 365 2021-09-03 06:46:15,152 bmiPcrglobwb INFO Shape of maps is (108, 126) 2021-09-03 06:46:15,152 bmiPcrglobwb INFO PCRGlobWB: initialize_model, source dir