MACCS processing failed

Hi,

I have in the system 3 sites and the system is correctly downloading Sentinel imagery. But when I look into the downloader history table there is 888 downloaded products, but 842 of them is in status 6 - Processing_failed. The rest is processed and is visible in the user interface. Coud you pls. help me what is wrong, what exact logs do you need to reslove where is the problem.

Thank you in advance

Jan Semrad

Did you have a look here

Yes but this is not the case, there are definetly images with very low cloud coverage, and they was not processed.
E.g.: this one S2B_MSIL2A_20190228T101029_N0207_R022_T33UVR_20190228T135024.SAFE with almost zero cloud coverage.

There is not in the failed product directory any EEF file only two log files see attachement.demmaccs.log (6.4 KB)
demmaccs_33UVR.log (8.9 KB)

What could be the problem? Do you need an other logs?

Check this Indices are not in ascending order - MSK_DETFOO

Obviously MACCS is having issues with S2 data since October 2018.

Is it working for earlier images?

Hello,

I think this is not the case - we are talking about imagery from this season (started in front end since 15th February and there is about 50 images processed properly (including some pictures from end of February). And of course it was working in last season (before Octobre 2018) but with different site areas and definitions.

Hello Jan,

If I am understanding correctly, you are trying products for 2019. As explained in the previous post MACCS processing failed, starting from October 2018 ESA introduced some changes in the product structure that MACCS cannot deal with (and it seems that CNES, the ones that are developing MACCS, will not do something to this too soon). Please note that not all the products since then have this problem as you might get sometimes also valid products.
As we are facing the same problem on our platforms, we are working now to integrate MAJA into the Sen2Agri system and a new version of the system will be published during April.
MAJA seem to handle correctly the issue but unfortunately the structure of the products differ.
As we are not the developers of MACCS in this moment we cannot come with another solution to solve your problem on this.
Anyway, in order to be sure that you are not facing another issue (disk space or others) you can do the following test:

  • go to the /mnt/archive/maccs_def/…/PRODUCT_NAME/ (a product that has no EEF file inside and you assume that it should have been processed OK, with no clouds)
  • check the log file with the name of the tile (ex. demmaccs_XXYYYY.log)
  • copy the command /usr/share/sen2agri/sen2agri-demmaccs/demmaccs.py --srtm /mnt/archive/srtm --swbd /mnt/archive/swbd --processes-number-dem 1 --processes-number-maccs 1 --gipp-dir /mnt/archive/gipp/ --working-dir /mnt/archive/demmaccs_tmp/ --maccs-launcher /opt/maccs/core/5.1/bin/maccs --delete-temp True /mnt/archive/dwn_def …
  • (that should be at the beginning of the file)
  • log in as user sen2agri-service (sudo su -l sen2agri-service)
  • execute the copied command
  • In the console you should see the detailed logs that might help you identify the problem.
    (Please check also this link to see if you get similar errors: https://forum.step.esa.int/t/indices-are-not-in-ascending-order-in-msk-detfoo/12812)

Best regards,
Cosmin

Hi @cudroiu,
Ok, I got the issue, but then what’s next? is not possible to try to ‘reprocess’ those ‘processing failed’ images?

Hello,

It will not be possible to produce a “processed” product after reprocessing those products with status “processing_failed”. The Sen2Agri team is currently validating/testing the integration of MAJA, the new atmospheric correction replacing MACCS.