[Opm] Parallel output in OPM Flow

Atgeirr Rasmussen Atgeirr.Rasmussen at sintef.no
Fri May 7 05:53:14 UTC 2021


Hi all,

I will try to answer the questions to the best of my knowledge. I will also ask someone from INRIA to joint this list, to be able to answer directly.


> Nice to hear that there is some work on the output layer. I just have  some questions.
> Can you please tell us what open source license Damaris uses. I could  not find any information on the website

It is LGPLv3+. It says in the source files (a LICENCE file should be added to the repo).

> Can you point us to a publication where it is used for reservoir simulation? That would be interesting.

I am not sure where this is published, I will ask.

>>So far, we see three important tasks related to this:
>> - Refactoring in OPM Flow to enable the integration of Damaris. From what we have seen so far, any necessary refactoring will be improving the Flow code, and make it more flexible.
>> - Integrating the Damaris library. We will ensure that this is an optional dependency, to minimize developer workload and hassle.

>What does optional mean here? If there is no Damaris we fall back to our
>current way and only if it is there we get hdf5/parallel output? Or are
>there plans to support hdf5 without Damaris (which seems to be a lot
>more than just output: data processing, in situ visualization,...)? That
>would be cool.

Optional means that without it, you get the existing output and no new dependencies, with it you will have a runtime option available for parallel hdf5 output. We could add HDF5 support without Damaris, certainly the code that must be written anyway to write out the actual output should only depend on HDF5, and be able to run with or without Damaris. The potential for in situ processing and visualization is certainly one of the aspects we will explore in ACROSS.

Atgeirr


More information about the Opm mailing list