GEOS-5

Running the GEOS-5 SBU Benchmark: Difference between revisions

← Older edit
 
(19 intermediate revisions by the same user not shown)
Line 1: Line 1:
==Build and install the model==
==Build and install the model==


First, untar the model tarball (in $NOBACKUP!!!):
First, untar the model tarball in <tt>nobackup</tt> or <tt>swdev</tt>; the model alone will break the <tt>home</tt> quota:


  $ tar xf Heracles-UNSTABLE-MPT-Benchmark.2017Feb13.tar.gz
  $ tar xf GEOSadas-5_16_5-Benchmark.tar.gz


Next, set up ESMADIR:
Next, set up ESMADIR:


  $ setenv ESMADIR <directory-to>/Heracles-UNSTABLE-MPT-Benchmark/GEOSagcm
  $ setenv ESMADIR <directory-to>/GEOSadas-5_16_5-Benchmark/GEOSadas


it is just below the src/ directory.
it is just below the src/ directory.
Line 84: Line 84:
  All 22 packages compiled successfully.
  All 22 packages compiled successfully.


In case of errors, gmh summarizes exactly where it happens by indicating the package where it occured. Caveat: it does not work in parallel (output is scrambled). So, if the parallel build fails, rerun it sequentially (it will go quickly and die in the same place) and run gmh on the output for a summary.
In case of errors, gmh summarizes exactly where it happens by indicating the package where it occurred. Caveat: it does not work in parallel (output is scrambled). So, if the parallel build fails, rerun it sequentially (it will go quickly and die in the same place) and run gmh on the output for a summary.
<!--
<!--
===Advanced features===
===Advanced features===
Line 106: Line 106:
These effectively let you change whatever you want - useful for debugging, etc. For example, you can set your timers in ~/.esma_base.mk.
These effectively let you change whatever you want - useful for debugging, etc. For example, you can set your timers in ~/.esma_base.mk.


-->
 
==Run the model==
==Test Build with One-Day Run==


To make sure all works, we will first try setting up a simple one-day experiment.
To make sure all works, we will first try setting up a simple one-day experiment.


===Setting up a one-day experiment===
===Setting up a one-day experiment===
Line 116: Line 117:


  $ cd $ESMADIR/src/Applications/GEOSgcm_App
  $ cd $ESMADIR/src/Applications/GEOSgcm_App
$ echo $NOBACKUP > ~/.HOMDIRroot
$ echo $NOBACKUP > ~/.EXPDIRroot
These echos set up some defaults. While they both don't have to be in the same location, it's highly recommended they are and the use of the script below assumes they are. Also, make sure they are in a nobackup directory.


Now you can run <tt>create_expt.py</tt>:
Now you can run <tt>create_expt.py</tt>:


  <nowiki>$ ~mathomp4/bin/create_exp.py -h
  <nowiki>$ ~mathomp4/bin/create_exp.py -h
usage: create_expt.py [-h] [-v] [-q] [--expdsc EXPDSC]
usage: create_expt.py [-h] [-v] [-q] [--expdsc EXPDSC] [--expdir EXPDIR]
                       [--horz {a,b,c,d,e,c12,c24,c48,c90,c180,c360,c720,c1440,c2880}]
                       [--horz {a,b,c,d,e,c12,c24,c48,c90,c180,c360,c720,c1440,c2880}]
                       [--vert {72,132}] [--ocean {o1,o2,o3}] [--land {1,2}]
                       [--vert {72,132}] [--ocean {o1,o2,o3}] [--land {1,2}]
Line 144: Line 141:
   -q, --quiet          Quietly Setup Experiment (no printing)
   -q, --quiet          Quietly Setup Experiment (no printing)
   --expdsc EXPDSC      Experiment Description (Default: same as expid)
   --expdsc EXPDSC      Experiment Description (Default: same as expid)
  --expdir EXPDIR      Experiment Directory Root *NOT CONTAINING EXPID*
                        (Default is what is in ~/.EXPDIRroot:
                        /discover/nobackup/mathomp4 )
   --horz {a,b,c,d,e,c12,c24,c48,c90,c180,c360,c720,c1440,c2880}
   --horz {a,b,c,d,e,c12,c24,c48,c90,c180,c360,c720,c1440,c2880}
                         Horizontal Resolution (Default: c48 on clusters, c12
                         Horizontal Resolution (Default: c48 on clusters, c12
Line 160: Line 160:
   --gpu                Setup Experiment to use GPUs</nowiki>
   --gpu                Setup Experiment to use GPUs</nowiki>


This is a script that attempts to ease setting up a GEOS-5 AGCM run. It has some "smarts" in that it will detect from an experiment ID some information, but you can be explicit (to the point you can specify impossible experiments too).
This is a script that attempts to ease setting up a GEOS-5 AGCM run. It has some "smarts" in that it will detect from an experiment ID some information, but you can be explicit (to the point you can specify impossible experiments too).  


To actually create the experiment, run the script:
Your first choice is where you'd like to run the experiment. $NOBACKUP or other disks with space are vital. For example, if you'd like to run the experiment in /discover/nobackup/username/experiment-name, then run the script with --expdir /discover/nobackup/username. To actually create the experiment, run the script and choose a C48 horizontal resolution:


  $ ~mathomp4/bin/create_expt.py test-1day-expt --horz c48 --account <ACCOUNTID>
  $ ~mathomp4/bin/create_expt.py test-1day-expt --horz c48 --account <ACCOUNTID> --expdir /discover/nobackup/username
  Horizontal resolution c48 passed in
  Horizontal resolution c48 passed in
  Using c48 horizontal resolution  
  Using c48 horizontal resolution  
Line 177: Line 177:
   Running gcm_setup...done.
   Running gcm_setup...done.
   
   
  Experiment is located in directory: /discover/nobackup/mathomp4/test-1day-expt
  Experiment is located in directory: /discover/nobackup/username/test-1day-expt


If you don't pass in an account-id, you'll get the default of g0620 (the developer's account).
If you don't pass in an account-id, you'll get the default of g0620 (the developer's account).
Line 185: Line 185:
To run a test one-day run, go to the experiment directory:
To run a test one-day run, go to the experiment directory:


  (1617) $ cd /discover/nobackup/mathomp4/test-1day-expt                                                       [discover08:...Applications/GEOSgcm_App 2:47pm]
  <nowiki>$ cd /discover/nobackup/mathomp4/test-1day-expt
total 96M
total 96M
-rwxr-xr-x 1 mathomp4 g0620 96M Feb 13 11:11 GEOSgcm.x*
-rwxr-xr-x 1 mathomp4 g0620 96M Feb 13 11:11 GEOSgcm.x*
Line 200: Line 200:
drwxr-xr-x 2 mathomp4 g0620 512 Feb 13 14:47 forecasts/
drwxr-xr-x 2 mathomp4 g0620 512 Feb 13 14:47 forecasts/
drwxr-xr-x 2 mathomp4 g0620 16K Feb 13 14:47 RC/
drwxr-xr-x 2 mathomp4 g0620 16K Feb 13 14:47 RC/
Directory: /discover/nobackup/mathomp4/test-1day-expt
Directory: /discover/nobackup/mathomp4/test-1day-expt</nowiki>
 
Now you'll run another script called <tt>makeoneday.bash</tt> that edits some .rc files, edits run scripts, as well as linking in some example restarts and sets up the model to stop after one day (with no options). Note: these restarts are write-protected, so if you want to run a multi-year run, don't use this script!
 
<nowiki>$ ~mathomp4/bin/makeoneday.bash
Using Heracles-5_0 directories
Making one-day experiment
 
Restoring AGCM.rc.save to AGCM.rc...
Copying AGCM.rc to AGCM.rc.save...
Restoring CAP.rc.save to CAP.rc...
Copying CAP.rc to CAP.rc.save...
Restoring HISTORY.rc.save to HISTORY.rc...
Copying HISTORY.rc to HISTORY.rc.save...
Restoring gcm_run.j.save to gcm_run.j...
Copying gcm_run.j to gcm_run.j.save...
Restoring regress/gcm_regress.j.save to regress/gcm_regress.j...
Copying regress/gcm_regress.j to regress/gcm_regress.j.save...
Restoring RC/GEOS_ChemGridComp.rc.save to RC/GEOS_ChemGridComp.rc...
Copying RC/GEOS_ChemGridComp.rc to RC/GEOS_ChemGridComp.rc.save...
DYN_INTERNAL_RESTART_TYPE not found. Assuming NC4
Found fvcore_internal_rst. Assuming you have needed restarts!
Changes made to CAP.rc:
9,10c9,10
< JOB_SGMT:    00000015 000000
< NUM_SGMT:    20
---
> JOB_SGMT:    00000001 000000
> NUM_SGMT:    1
25c25
< MAPL_ENABLE_TIMERS: NO
---
> MAPL_ENABLE_TIMERS: YES
 
 
Changes made to gcm_run.j:
7c7
< #PBS -l walltime=12:00:00
---
> #PBS -l walltime=0:15:00
11a12
> #SBATCH --mail-type=ALL
260c261
< if($numrs == 0) then
---
> if($numrs == 1) then
433a435
> exit
</nowiki>
 
At this point, you should be able to <tt>sbatch gcm_run.j</tt> and the model should run a day.
 
-->
 
==Benchmark Run==
 
The full benchmark run is a run of 5-days using a portable version of the GEOS-5 boundary conditions. This will use space and cores. For effective benchmarking of I/O, it's recommended to run on less congested than nobackup.
 
===Learn to love tcsh===
 
One preliminary note is that GEOS-5 is, in many ways, a collection of csh/tcsh scripts. If things start going wrong, the answer can often be "change your shell to tcsh and try". Yes, it's not bash/fish/zsh, but it is what it is. I don't think it's entirely necessary for just this automated work, but it could happen.
 
===Setting up benchmark experiment===
 
Go into the model application directory and do a couple of preliminary commands:
 
$ cd $ESMADIR/src/Applications/GEOSgcm_App
 
These echos set up some defaults. While they both don't have to be in the same location, it's highly recommended they are and the use of the script below assumes they are. Also, make sure they are in a nobackup directory.
 
For the next few commands, you will need to know the location of the portable BCs directory used for this experiment, which is referred to below as <tt>$PORTBCS</tt>. On discover, a version will always be at:
 
/discover/nobackup/mathomp4/HugeBCs-H50
 
To create the experiment, run create_expt.py and choose a C720 horizontal resolution with climatological GOCART:
 
$ $PORTBCS/scripts/create_expt.py benchmark-GEOSadas-5_16-5-5day-c720 --horz c720 --ocean o3 --gocart C --account <ACCOUNTID> --expdir <root-for-experiment>
Found c720 horizontal resolution in experiment name
Using c720 horizontal resolution
Assuming default vertical resolution of 72
Using 72 vertical resolution
Ocean resolution of o3 passed in
Using o3 ocean resolution
Using climatological aerosols
  Running gcm_setup...done.
Experiment is located in directory: <root-for-experiment>/benchmark-GEOSadas-5_16-5-5day-c720
 
Again, if you don't pass in an account-id, you'll get the default of g0620 (the developer's account).
 
===Setup and Run Benchmark===
 
Now ''change to the experiment directory'' and run MakeSBUBench.bash which will set up the experiment:
 
$ $PORTBCS/scripts/MakeSBUBench.bash
 
The script sets the run up for 5 days using 5400 cores, and other flags are tripped to best emulate Ops.
 
'''NOTE''': This will also set the experiment to run in this SLURM environment:
 
#SBATCH --partition=preops
#SBATCH --qos=benchmark
 
This is how the script's developer can run 5400-core jobs. Others might have different partition/qos to submit to. Please edit these before sbatch submission to a <tt>partition/qos</tt> that you have access to that can accept a 5400-core job.
 
Finally, submit the job:
 
$ sbatch gcm_run.j




Retrieved from "https://geos5.org/wiki/index.php?title=Running_the_GEOS-5_SBU_Benchmark"