Measures hdf5 serial performance.
h5perf_serial: Version 1.8.7 usage: h5perf_serial [OPTIONS]
OPTIONS
-h
Print an usage message and exit
-A AL
Which APIs to test [default: all of them]
-c SL
Selects chunked storage and defines chunks dimensions and sizes [default: Off]
-e SL
Dimensions and sizes of dataset [default: 100,200]
-i N
Number of iterations to perform [default: 1]
-r NL
Dimension access order (see below for description) [default: 1,2]
-t
Selects extendable dimensions for HDF5 dataset [default: Off]
-v VFD
Selects file driver for HDF5 access [default: sec2]
-w
Perform write tests, not the read tests [default: Off]
-x SL
Dimensions and sizes of the transfer buffer [default: 10,20]
N
- is an integer > 0.
S
- is a size specifier, an integer > 0 followed by a size indicator:
K - Kilobyte (1024) M - Megabyte (1048576) G - Gigabyte (1073741824)
Example: '37M' is 37 megabytes or 38797312 bytes
AL - is an API list. Valid values are:
hdf5 - HDF5 posix - POSIX
Example: -A posix,hdf5
NL - is list of integers (N) separated by commas.
Example: 1,2,3
SL - is list of size specifiers (S) separated by commas.
Example: 2K,2K,3K
The example defines an object (dataset, tranfer buffer) with three dimensions. Be aware that as the number of dimensions increases, the the total size of the object increases exponentially.
VFD
- is an HDF5 file driver specifier. Valid values are:
sec2, stdio, core, split, multi, family, direct
Dimension access order:
Data access starts at the cardinal origin of the dataset using the transfer buffer. The next access occurs on a dataset region next to the previous one. For a multidimensional dataset, there are several directions as to where to proceed. This can be specified in the dimension access order. For example, -r 1,2 states that the tool should traverse dimension 1 first, and then dimension 2.
Environment variables:
HDF5_NOCLEANUP
Do not remove data files if set [default remove]
HDF5_PREFIX
Data file prefix
usage: h5perf_serial [OPTIONS]
OPTIONS
-h
Print an usage message and exit
-A AL
Which APIs to test [default: all of them]
-c SL
Selects chunked storage and defines chunks dimensions and sizes [default: Off]
-e SL
Dimensions and sizes of dataset [default: 100,200]
-i N
Number of iterations to perform [default: 1]
-r NL
Dimension access order (see below for description) [default: 1,2]
-t
Selects extendable dimensions for HDF5 dataset [default: Off]
-v VFD
Selects file driver for HDF5 access [default: sec2]
-w
Perform write tests, not the read tests [default: Off]
-x SL
Dimensions and sizes of the transfer buffer [default: 10,20]
N
- is an integer > 0.
S
- is a size specifier, an integer > 0 followed by a size indicator:
K - Kilobyte (1024) M - Megabyte (1048576) G - Gigabyte (1073741824)
Example: '37M' is 37 megabytes or 38797312 bytes
AL - is an API list. Valid values are:
hdf5 - HDF5 posix - POSIX
Example: -A posix,hdf5
NL - is list of integers (N) separated by commas.
Example: 1,2,3
SL - is list of size specifiers (S) separated by commas.
Example: 2K,2K,3K
The example defines an object (dataset, tranfer buffer) with three dimensions. Be aware that as the number of dimensions increases, the the total size of the object increases exponentially.
VFD
- is an HDF5 file driver specifier. Valid values are:
sec2, stdio, core, split, multi, family, direct
Dimension access order:
Data access starts at the cardinal origin of the dataset using the transfer buffer. The next access occurs on a dataset region next to the previous one. For a multidimensional dataset, there are several directions as to where to proceed. This can be specified in the dimension access order. For example, -r 1,2 states that the tool should traverse dimension 1 first, and then dimension 2.
Environment variables:
HDF5_NOCLEANUP
Do not remove data files if set [default remove]
HDF5_PREFIX
Data file prefix
The full documentation for h5perf_serial: is maintained as a Texinfo manual. If the info and h5perf_serial: programs are properly installed at your site, the command
info h5perf_serial:
should give you access to the complete manual.