- Infos im HLRS Wiki sind nicht rechtsverbindlich und ohne Gewähr -
- Information contained in the HLRS Wiki is not legally binding and HLRS is not responsible for any damages that might result from its use -

Scalasca: Difference between revisions

From HLRS Platforms
Jump to navigationJump to search
No edit summary
 
(Add link to Score-P)
 
(8 intermediate revisions by 2 users not shown)
Line 1: Line 1:
{{Infobox software
{{Infobox software
| logo = [[Image:scalasca-logo.png]]
| description = Scalasca is an open-source toolset for the performance analyzes of parallel applications helping to identify optimization opportunities. It has been specifically designed for the use on large-scale systems including Cray XT/XE, but is also well-suited for small- and medium-scale HPC platforms. Scalasca supports an incremental performance-analysis procedure that integrates runtime summaries with in-depth studies of concurrent behavior via event tracing, adopting a strategy of successively refined measurement configurations. A distinctive feature is the ability to identify wait states that occur, for example, as a result of unevenly distributed workloads.
| description = Scalasca is an open-source toolset for the performance analyzes of parallel applications helping to identify optimization opportunities. It has been specifically designed for the use on large-scale systems including Cray XT/XE, but is also well-suited for small- and medium-scale HPC platforms. Scalasca supports an incremental performance-analysis procedure that integrates runtime summaries with in-depth studies of concurrent behavior via event tracing, adopting a strategy of successively refined measurement configurations. A distinctive feature is the ability to identify wait states that occur, for example, as a result of unevenly distributed workloads.
| developer              = JSC
| developer              = JSC
Line 5: Line 6:
| category                  = [[:Category:Performance Analyzer | Performance Analyzer]]
| category                  = [[:Category:Performance Analyzer | Performance Analyzer]]
| license                = New BSD
| license                = New BSD
| website                = [http://www.fz-juelich.de/jsc/scalasca/ Scalasca homepage]  
| website                = [http://www.scalasca.org Scalasca homepage]  
}}
}}


== Usage ==
== Usage ==


To use Scalsca first load the respective software module:
To use Scalasca start by loading the respective software modules:
{{Command|command =
<nowiki /># on HAWK:
module load scalasca scorep cube


{{Command|command =
<nowiki /># on Vulcan
module load performance/scalasca
module load performance/scalasca performance/scorep
}}
}}


First instrument your application using the scin (scalasca -instrument) command and the MPI compiler
Then, instrument your application using [[Score-P]] by prefixing all MPI compiler and linker commands with the ''scorep'' command:
{{Command|command =  
{{Command|command =  
skin mpicc myapp.c -o myapp
scorep mpicc myapp.c -o myapp
}}
}}
Then analyze your application using the scan (scalasca -analyze) command and ''mpirun''
 
Next analyze your application using the ''scalasca -analyze'' command with your usual ''mpirun'' command:
{{Command|command=
{{Command|command=
scan mpirun -np ${NP} myapp
export MPI_SHEPHERD=true # this is mandatory when running with mpt on HAWK<br>
scalasca -analyze mpirun -n ${NP} myapp
}}
}}
The last step is to explore the generated reports by rerunning your application using the square (scalasca -examine) command:
After the run there will be a folder prefixed with "scorep_*" containing the messurment results.
{{Warning|text=The scalasca -analyze command evaluates the provided MPI command line and will try to determine the used MPI implementation, number of MPI ranks (based on '-n' option), additional MPI options, the application name, and the application's arguments. If you encounter problems at this point, please consult the scalasca documentation for details.}}
 
The last step is to explore the generated reports by using the ''scalasca -examine'' command, with the previously created "scorep_*" folder as the experiment name:
{{Command|command=
{{Command|command=
square mpirun -np ${NP} myapp
scalasca -examine <experiment_name>
}}
This will open Cube with the generated report automatically if you have enabled X forwarding.
You can also inspect the generated ''*.cubex'' report files, manually starting the Cube GUI:
{{Command|command=
cube <file>.cubex
}}
}}


To view the final results you then can use the Cube3 GUI
=== Usage on Hawk ===
{{Command|command=
{{Warning|text=
cube3
It is necessary to set {{command|command=export MPI_SHEPHERD=true}} for the ''scalasca -analyze'' command.
}}
}}


Line 38: Line 52:


== External links ==
== External links ==
* [http://www.fz-juelich.de/jsc/scalasca/ Scalasca homepage]
* [http://www.scalasca.org Scalasca homepage]


[[Category:Performance Analyzer]]
[[Category:Performance Analyzer]]

Latest revision as of 13:27, 29 April 2020

Scalasca is an open-source toolset for the performance analyzes of parallel applications helping to identify optimization opportunities. It has been specifically designed for the use on large-scale systems including Cray XT/XE, but is also well-suited for small- and medium-scale HPC platforms. Scalasca supports an incremental performance-analysis procedure that integrates runtime summaries with in-depth studies of concurrent behavior via event tracing, adopting a strategy of successively refined measurement configurations. A distinctive feature is the ability to identify wait states that occur, for example, as a result of unevenly distributed workloads.
Scalasca-logo.png
Developer: JSC
Platforms:
Category: Performance Analyzer
License: New BSD
Website: Scalasca homepage


Usage

To use Scalasca start by loading the respective software modules:

# on HAWK:

module load scalasca scorep cube

# on Vulcan

module load performance/scalasca performance/scorep


Then, instrument your application using Score-P by prefixing all MPI compiler and linker commands with the scorep command:

scorep mpicc myapp.c -o myapp


Next analyze your application using the scalasca -analyze command with your usual mpirun command:

export MPI_SHEPHERD=true # this is mandatory when running with mpt on HAWK
scalasca -analyze mpirun -n ${NP} myapp

After the run there will be a folder prefixed with "scorep_*" containing the messurment results.

Warning: The scalasca -analyze command evaluates the provided MPI command line and will try to determine the used MPI implementation, number of MPI ranks (based on '-n' option), additional MPI options, the application name, and the application's arguments. If you encounter problems at this point, please consult the scalasca documentation for details.


The last step is to explore the generated reports by using the scalasca -examine command, with the previously created "scorep_*" folder as the experiment name:

scalasca -examine <experiment_name>

This will open Cube with the generated report automatically if you have enabled X forwarding. You can also inspect the generated *.cubex report files, manually starting the Cube GUI:

cube <file>.cubex


Usage on Hawk

Warning: It is necessary to set
export MPI_SHEPHERD=true
for the scalasca -analyze command.


See also

External links