Hi Dan,
Thanks for your response! I am not running it in parallel. It was my understanding that to run in parallel I would have to be analysing multiple frames. In this case I am just attempting to run nmode for a single frame. I recently restarted the job on a supercomputer cluster with a large memory request (32gb instead of 4gb). Thus far, the job has run without an issue longer than my previous attempt which was killed for excessive memory usage (when I had only requested 4gb of memory). Hopefully that job will complete successfully, but I am curious why my job seems to be requiring more memory than your estimate.
Best,
Nick
________________________________________
From: Daniel Roe <daniel.r.roe.gmail.com>
Sent: Wednesday, March 23, 2016 11:54 AM
To: AMBER Mailing List
Subject: Re: [AMBER] Normal Mode Job Allocation Failure
Hi,
Are you by any chance running mmpbsa.py in parallel? By my calculation
4373 atoms should require about 2 GB of memory for nmode; however, if
you are running in parallel that's 2 GB for each thread (not counting
the memory required for other things), which can get out of hand at
higher thread counts.
-Dan
On Mon, Mar 21, 2016 at 9:18 AM, Cockroft, Nicholas T.
<cockroft.1.buckeyemail.osu.edu> wrote:
> Hi,
>
>
> I have been running single frame nmode analysis jobs on a series of complexes and have encountered an issue with my largest complex. The receptor is 562 residues/4373 atoms in size. The error I get upon job failure is:
>
>
>
> FATAL: allocation failure in vector()
>
> File "/usr/local/amber14/bin/MMPBSA.py", line 104, in <module>
> app.run_mmpbsa()
> File "/usr/local/amber14/lib/python2.6/site-packages/MMPBSA_mods/main.py", line 218, in run_mmpbsa
> self.calc_list.run(rank, self.stdout)
> File "/usr/local/amber14/lib/python2.6/site-packages/MMPBSA_mods/calculation.py", line 82, in run
> calc.run(rank, stdout=stdout, stderr=stderr)
> File "/usr/local/amber14/lib/python2.6/site-packages/MMPBSA_mods/calculation.py", line 321, in run
> Calculation.run(self, rank, stdout=self.output % rank)
> File "/usr/local/amber14/lib/python2.6/site-packages/MMPBSA_mods/calculation.py", line 157, in run
> self.prmtop))
> CalcError: /usr/local/amber14/bin/mmpbsa_py_nabnmode failed with prmtop ../1ER_4IH7_dry.prmtop!
> Exiting. All files have been retained.
>
>
>
> It appears it is failing due to insufficient memory. I previously ran this job on a supercomputer cluster and it was killed due to excessive memory usage. It had been allocated 4gb of memory for that job. I then re-ran it on my local workstation with at least 8 gb of memory available for the calculation (and got the error above). What steps do I need to take to reduce the memory requirement to get the calculation to finish successfully?
>
>
> Thanks,
>
> Nick
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
--
-------------------------
Daniel R. Roe, PhD
Department of Medicinal Chemistry
University of Utah
30 South 2000 East, Room 307
Salt Lake City, UT 84112-5820
http://home.chpc.utah.edu/~cheatham/
(801) 587-9652
(801) 585-6208 (Fax)
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Mar 23 2016 - 11:30:04 PDT
