Re: [AMBER] recovering amber binaries

From: Pratul Agarwal <>
Date: Tue, 22 Jan 2019 13:41:03 +0000

Probably best to recompile AMBER for the long term.

In the meantime here is a way to locate binaries:
find /path/to/lost+found -type f -print -exec ldd {} \;

This will give you a list of dynamically linked libraries needed for files that are binaries
For pmemd.MPI (and other MPI programs) you will see (if you compiled using MPICH) in the list of libraries
For pmemd.cuda (and other CUDA programs) you will see etc.

After some inspection and guess work you should be able to locate the binary files you want. Once located you can execute these binaries in a temporary folder -- the header/output/error will give you some more information. But if your libraries were also moved you will have additional issues while trying to execute these binary files.

Good luck.

Pratul K. Agarwal, Ph.D.
(Editorial Board Member: PLoS ONE, Microbial Cell Factories)

On 1/22/2019 7:12 AM, Andreas Tosstorff wrote:

Hi all,

today I received an email that said "Bad news". Turns out that a lot of
data on our cluster was lost.

Does anybody have experience with data recovery?

We may be able to recover files, however the folder structure and file
names are lost. Text files could be identified by string search. I was
wondering if there may be a way to identify amber binary files in a
similar way?

Any help is highly appreciated!

Kind regards,


  PIPPI conference

M.Sc. Andreas Tosstorff
Lehrstuhl für Pharmazeutische Technologie und Biopharmazie
Department Pharmazie
LMU München
Butenandtstr. 5-13 ( Haus B)
81377 München

Tel. +49 (0) 89 2180 77059
Email :<>
Homepage :

AMBER mailing list<>

AMBER mailing list
Received on Tue Jan 22 2019 - 06:00:02 PST
Custom Search