Singularity image for distribution

[nschiraldi]

I have been working on building a singularity image for HydroBlocks as some of our researchers are interested in using the application in our HPC environment. Is it of any interest to the developers of HydroBlocks to have a singularity (or docker) image to encourage reproducibility and transparency of the software?

The .def file is here: https://github.com/nschiraldi/singularity/blob/master/Singularity.HydroBlocks and it is being hosted on singularity-hub. It is not yet operational, but I figured I'd start the thread in case anyone would like to contribute.

Current usage only supports the "preprocessing" stage. For example:

$ singularity pull shub://nschiraldi/singularity:HydroBlocks
$ singularity run --app preprocessing hydroblocks metadata.json

[chaneyn]

This sounds very interesting. I have always been intrigued by taking the singularity/docker route to make HydroBlocks available to the larger community. Excited to see how things turn out. Nate

…

On Nov 14, 2019, at 9:25 AM, nschiraldi ***@***.***> wrote: I have been working on building a singularity image for HydroBlocks as some of our researchers are interested in using the application in our HPC environment. Is it of any interest to the developers of HydroBlocks to have a singularity (or docker) image to encourage reproducibility and transparency of the software? The .def file is here: https://github.com/nschiraldi/singularity/blob/master/Singularity.HydroBlocks <https://github.com/nschiraldi/singularity/blob/master/Singularity.HydroBlocks> and it is being hosted on singularity-hub. It is not yet operational, but I figured I'd start the thread in case anyone would like to contribute. Current usage only supports the "preprocessing" stage. For example: $ singularity pull shub://nschiraldi/singularity:HydroBlocks $ singularity run --app preprocessing hydroblocks metadata.json — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#6?email_source=notifications&email_token=AAUJXZUBDE2QHG3FMFXEUUTQTVNWXA5CNFSM4JNMMVA2YY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HZKWZBQ>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAUJXZT4FT43HZ3TVUBGND3QTVNWXANCNFSM4JNMMVAQ>.

[nschiraldi]

Just a quick update, had some more time to revisit this and I have the "preprocessing" working on the HB_sample data. Working on some of the other tasks this week. Also looking at getting the environment optimized with the intel channel in conda which has some big impacts on performance from our tests.