Commit 4d085c2d authored by Matthieu Dorier's avatar Matthieu Dorier
Browse files


parent bcfd3c21
......@@ -25,15 +25,17 @@ Darshan must be installed with bzip2 support, and with the --enable-shared
build option. Assuming the path to Darshan is "path/to/darshan", install
Darshan-Ruby can be done as follows:
> gem install darshan-ruby -- --with-darshan-dir=path/to/darshan
gem install darshan-ruby -- --with-darshan-dir=path/to/darshan
If you use [RVM]( to manage your packages:
> rvm gemset create darshan
> rvm gemset use darshan
> gem install darshan-ruby -- --with-darshan-dir=path/to/darshan
rvm gemset create darshan
rvm gemset use darshan
gem install darshan-ruby -- --with-darshan-dir=path/to/darshan
......@@ -43,11 +45,15 @@ Quarshan is a command-line tool to query darshan files.
Once Darshan-Ruby is installed, you can call Quarshan as follows:
> quarshan -o COUNTER file.darshan
quarshan -o COUNTER file.darshan
For example:
> quarshan -o POSIX_BYTES_WRITTEN ior1.darshan
quarshan -o POSIX_BYTES_WRITTEN ior1.darshan
Will output
......@@ -58,6 +64,45 @@ Will output
/home/shane/software/darshan/testFile.00000003 1048576
You can obtain a header using the **-v** option:
> quarshan -o POSIX_BYTES_WRITTEN ior1.darshan -v
/home/shane/software/darshan/testFile.00000000 1048576
/home/shane/software/darshan/testFile.00000001 1048576
/home/shane/software/darshan/testFile.00000002 1048576
/home/shane/software/darshan/testFile.00000003 1048576
You can ask for multiple counters by separating them by commas (without spaces).
These counters must all be part of the same module (here POSIX, for example):
> quarshan -o POSIX_BYTES_WRITTEN,POSIX_SIZE_WRITE_100K_1M ior1.darshan -v
/home/shane/software/darshan/testFile.00000000 1048576 4
/home/shane/software/darshan/testFile.00000001 1048576 4
/home/shane/software/darshan/testFile.00000002 1048576 4
/home/shane/software/darshan/testFile.00000003 1048576 4
You can provide reduction operations to directly compute statistics using **-r**:
quarshan -o POSIX_BYTES_WRITTEN,POSIX_SIZE_WRITE_100K_1M -r avg,min ior1.darshan -v
4 1048576.0 4
Instead of printing the name of the record on the first column, the number of
records will be printed.
Available reduction operators are: min (minimum), max (maximum), avg (average)
var (variance), std (standard deviation), med (median), sum (sum).
When providing reduction operations, if you provide only one reduction
operation, it will be applied to all requested counters. If you provide multiple
reduction operations, you must make sure to provide the same number as the number
of counters.
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment