Skip to content

Commit

Permalink
docs: refresh for v0.10
Browse files Browse the repository at this point in the history
Signed-off-by: Adam Moody <[email protected]>
  • Loading branch information
adammoody committed Jan 27, 2020
1 parent 355d7b7 commit b38043e
Show file tree
Hide file tree
Showing 14 changed files with 70 additions and 56 deletions.
2 changes: 1 addition & 1 deletion man/dbcast.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DBCAST" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DBCAST" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dbcast \- distributed broadcast
.
Expand Down
2 changes: 1 addition & 1 deletion man/dbz2.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DBZ2" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DBZ2" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dbz2 \- distributed bz2 compression
.
Expand Down
2 changes: 1 addition & 1 deletion man/dchmod.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DCHMOD" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DCHMOD" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dchmod \- distributed tool to set permissions and group
.
Expand Down
2 changes: 1 addition & 1 deletion man/dcmp.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DCMP" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DCMP" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dcmp \- distributed compare
.
Expand Down
2 changes: 1 addition & 1 deletion man/dcp.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DCP" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DCP" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dcp \- distributed copy
.
Expand Down
2 changes: 1 addition & 1 deletion man/ddup.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DDUP" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DDUP" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
ddup \- report files with identical content
.
Expand Down
2 changes: 1 addition & 1 deletion man/dfind.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DFIND" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DFIND" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dfind \- distributed file filtering
.
Expand Down
2 changes: 1 addition & 1 deletion man/dreln.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DRELN" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DRELN" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dreln \- distributed relink
.
Expand Down
2 changes: 1 addition & 1 deletion man/drm.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DRM" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DRM" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
drm \- distributed remove
.
Expand Down
2 changes: 1 addition & 1 deletion man/dstripe.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DSTRIPE" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DSTRIPE" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dstripe \- restripe files on underlying storage
.
Expand Down
2 changes: 1 addition & 1 deletion man/dsync.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DSYNC" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DSYNC" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dsync \- synchronize directory trees
.
Expand Down
9 changes: 8 additions & 1 deletion man/dwalk.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "DWALK" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "DWALK" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
dwalk \- distributed walk and list
.
Expand Down Expand Up @@ -103,6 +103,13 @@ Print files to the screen.
.UNINDENT
.INDENT 0.0
.TP
.B \-\-progress N
Print progress message to stdout approximately every N seconds.
The number of seconds must be a non\-negative integer.
A value of 0 disables progress messages.
.UNINDENT
.INDENT 0.0
.TP
.B \-v, \-\-verbose
Run in verbose mode.
.UNINDENT
Expand Down
93 changes: 50 additions & 43 deletions man/mpifileutils.1
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
.\" Man page generated from reStructuredText.
.
.TH "MPIFILEUTILS" "1" "Dec 28, 2019" "0.10" "mpiFileUtils"
.TH "MPIFILEUTILS" "1" "Jan 27, 2020" "0.10" "mpiFileUtils"
.SH NAME
mpifileutils \- mpiFileUtils Documentation
.
Expand Down Expand Up @@ -42,6 +42,41 @@ jobs like copy, remove, and compare for such datasets, providing speedups of up
to 50x. The libmfu library simplifies the creation of new tools
and it can be called directly from within HPC applications.
.SH USER GUIDE
.SS Project Design Principles
.sp
The following principles drive design decisions in the project.
.SS Scale
.sp
The library and tools should be designed such that running with more processes
increases performance, provided there are sufficient data and parallelism
available in the underlying file systems. The design of the tool should not
impose performance scalability bottlenecks.
.SS Performance
.sp
While it is tempting to mimic the interface, behavior, and file formats of
familiar tools like cp, rm, and tar, when forced with a choice between
compatibility and performance, mpiFileUtils chooses performance. For example,
if an archive file format requires serialization that inhibits parallel
performance, mpiFileUtils will opt to define a new file format that enables
parallelism rather than being constrained to existing formats. Similarly,
options in the tool command line interface may have different semantics from
familiar tools in cases where performance is improved. Thus, one should be
careful to learn the options of each tool.
.SS Portability
.sp
The tools are intended to support common file systems used in HPC centers, like
Lustre, GPFS, and NFS. Additionally, methods in the library should be portable
and efficient across multiple file systems. Tool and library users can rely on
mpiFileUtils to provide portable and performant implementations.
.SS Composability
.sp
While the tools do not support chaining with Unix pipes, they do support
interoperability through input and output files. One tool may process a dataset
and generate an output file that another tool can read as input, e.g., to walk
a directory tree with one tool, filter the list of file names with another, and
perhaps delete a subset of matching files with a third. Additionally, when logic
is deemed to be useful across multiple tools or is anticipated to be useful in
future tools or applications, it should be provided in the common library.
.SS Utilities
.sp
The tools in mpiFileUtils are MPI applications. They must be launched
Expand Down Expand Up @@ -225,12 +260,12 @@ installdir=\(gapwd\(ga/install

mkdir deps
cd deps
wget https://github.com/hpc/libcircle/releases/download/0.2.1\-rc.1/libcircle\-0.2.1\-rc.1.tar.gz
wget https://github.com/hpc/libcircle/releases/download/v0.3/libcircle\-0.3.0.tar.gz
wget https://github.com/llnl/lwgrp/releases/download/v1.0.2/lwgrp\-1.0.2.tar.gz
wget https://github.com/llnl/dtcmp/releases/download/v1.1.0/dtcmp\-1.1.0.tar.gz

tar \-zxf libcircle\-0.2.1\-rc.1.tar.gz
cd libcircle\-0.2.1\-rc.1
tar \-zxf libcircle\-0.3.0.tar.gz
cd libcircle\-0.3.0
./configure \-\-prefix=$installdir
make install
cd ..
Expand All @@ -256,17 +291,17 @@ To build on PowerPC, one may need to add \fB\-\-build=powerpc64le\-redhat\-linux
to the configure commands.
.sp
Assuming the dependencies have been placed in
an \fIinstall\fP directory as shown above, build mpiFileUtils from a release like v0.9:
an \fIinstall\fP directory as shown above, build mpiFileUtils from a release like v0.10:
.INDENT 0.0
.INDENT 3.5
.sp
.nf
.ft C
wget https://github.com/hpc/mpifileutils/archive/v0.9.tar.gz
tar \-zxf v0.9.tar.gz
wget https://github.com/hpc/mpifileutils/archive/v0.10.tar.gz
tar \-zxf v0.10.tar.gz
mkdir build install
cd build
cmake ../mpifileutils\-0.9 \e
cmake ../mpifileutils\-0.10 \e
\-DWITH_DTCMP_PREFIX=../install \e
\-DWITH_LibCircle_PREFIX=../install \e
\-DCMAKE_INSTALL_PREFIX=../install
Expand Down Expand Up @@ -340,41 +375,6 @@ cmake ../mpifileutils
.sp
The other way to use spack is to create a “view” to the installed dependencies.
Details on this are coming soon.
.SS Project Design Principles
.sp
The following principles drive design decisions in the project.
.SS Scale
.sp
The library and tools should be designed such that running with more processes
increases performance, provided there are sufficient data and parallelism
available in the underlying file systems. The design of the tool should not
impose performance scalability bottlenecks.
.SS Performance
.sp
While it is tempting to mimic the interface, behavior, and file formats of
familiar tools like cp, rm, and tar, when forced with a choice between
compatibility and performance, mpiFileUtils chooses performance. For example,
if an archive file format requires serialization that inhibits parallel
performance, mpiFileUtils will opt to define a new file format that enables
parallelism rather than being constrained to existing formats. Similarly,
options in the tool command line interface may have different semantics from
familiar tools in cases where performance is improved. Thus, one should be
careful to learn the options of each tool.
.SS Portability
.sp
The tools are intended to support common file systems used in HPC centers, like
Lustre, GPFS, and NFS. Additionally, methods in the library should be portable
and efficient across multiple file systems. Tool and library users can rely on
mpiFileUtils to provide portable and performant implementations.
.SS Composability
.sp
While the tools do not support chaining with Unix pipes, they do support
interoperability through input and output files. One tool may process a dataset
and generate an output file that another tool can read as input, e.g., to walk
a directory tree with one tool, filter the list of file names with another, and
perhaps delete a subset of matching files with a third. Additionally, when logic
is deemed to be useful across multiple tools or is anticipated to be useful in
future tools or applications, it should be provided in the common library.
.SH MAN PAGES
.SS dbcast
.SS SYNOPSIS
Expand Down Expand Up @@ -1819,6 +1819,13 @@ Print files to the screen.
.UNINDENT
.INDENT 0.0
.TP
.B \-\-progress N
Print progress message to stdout approximately every N seconds.
The number of seconds must be a non\-negative integer.
A value of 0 disables progress messages.
.UNINDENT
.INDENT 0.0
.TP
.B \-v, \-\-verbose
Run in verbose mode.
.UNINDENT
Expand Down
2 changes: 1 addition & 1 deletion mpifileutils.spec
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
Name: mpifileutils
Version: 0.9.1
Version: 0.10
Release: 1%{?dist}
Summary: File utilities designed for scalability and performance.

Expand Down

0 comments on commit b38043e

Please sign in to comment.