Problems with walkers MPI - plumed 2.8.0-dev with Gromacs 2019.6

[casalini...@gmail.com]

Dear Plumed users and developers,

I have an unexpected issue with the walkers_mpi option. 

I am using Gromacs 2019.6 patched with plumed 2.8.0-dev version on gpu. 

I am running WTMetaD simulations using the funnel module with three replicas, adopting the walkers_mpi option.

The first 50ns were fine, although I see that I have three COLVAR files but only one HILLS file, which contains the bias for the three replicas. I am using the -multidir flag and the three directories are called 0, 1 and 2; HILLS file is only in folder 0. 

When I tried to run additional 50 ns, the deposited bias was read only in directory 0, while it was ignored by directories 1 and 2. It can be clearly seen by the HILLS file, since the second and in the third replica the height of the hills went back to the initial value. 

Did I miss something? 

I thank you in advance for your support.

Best regards,

Tommaso

[Carlo Camilloni]

It is a common issue we need to simplify.

Anyway it is enough to read and write hills file from the same folder using 

FILE=../HILLS

In all plumed input files

Carlo

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/0193a186-94b7-4df1-a97b-e4879e5979e9n%40googlegroups.com.

[Billy Noonan]

Hi Carlo and Casalini,

Sorry to bring up an old thread, but any help with the following would be much appreciated...
https://groups.google.com/g/plumed-users/c/RiVveeYZbN8

Best Wishes,
Billy