Sen2Cor processing error for 3 specific Sentinel-2 products

[Fahdben]

For the three products below, the sen2cor atmospheric correction stops at approx. 80% of its progress.
Products:

• S2A_MSIL1C_20190519T105621_N0207_R094_T30SVH_20190519T112547
• S2A_MSIL1C_20190519T105621_N0207_R094_T30SWH_20190519T112547
• S2A_MSIL1C_20190728T105621_N0208_R094_T30SUH_20190728T130259
  The only message we get about this error is : [2021-08-31 11:27:10,607] [world-cereal] [ERROR ] Killed
  Some possible reasons for this bug and solutions:
• Insufficient RAM allocated to a pod for these products, usually 7 GB of RAM is enough but for the 3 products more is needed. @agrosu1978 tested this hypothesis and seems to fix this issue. But we can't reliably tell the max RAM needed for sen2cor to work every time
• Check the default sen2cor parameters used
• Compare L1C data from eodata and creodias finder

[savmickael]

Additional test to do: run the docker with ram limit set to 7 or 8 Gb

[savmickael]

In Sen2Cor 2.9 SUM a memory requirement of 4Gb is indicated but in the step forum, some users reported issue with memory:

• https://forum.step.esa.int/t/is-there-any-requirements-about-the-the-computer-memory-when-using-sen2cor-to-process-the-sentinel-2-data/10667
• https://forum.step.esa.int/t/memoryerror-sen2cor-for-10m-res/1430
• https://forum.step.esa.int/t/sen2cor-processing-not-all-tiles-under-vm/26799

[savmickael]

Add Validation plan this test with a ram limit