mpip
Table of Content
About
mpiP is a lightweight profiling library for MPI applications.
- Software was developed by LLNL (Lawrence Livermore National Laboratory).
- Collects only statistical information about MPI routines
- Captures and stores information local to each task (local memory and disk)
- Uses communication only at the end of the application to merge results from all tasks into one output file.
The MPI call statistical information provided includes:
- Percent of a task's time attributed to MPI calls
- Where each MPI call is made within the program (callsites)
- Top 20 callsites
- Callsite statistics (for all callsites)
Versions and Availability
▶ Display Softenv Keys for mpip on all clusters
| Machine | Version | Softenv Key |
|---|---|---|
| None Available | N/A | N/A |
▶ Softenv FAQ?
The information here is applicable to LSU HPC and LONI systems.
Shells
A user may choose between using /bin/bash and /bin/tcsh. Details about each shell follows.
/bin/bash
System resource file: /etc/profile
When one access the shell, the following user files are read in if they exist (in order):
- ~/.bash_profile (anything sent to STDOUT or STDERR will cause things like rsync to break)
- ~/.bashrc (interactive login only)
- ~/.profile
When a user logs out of an interactive session, the file ~/.bash_logout is executed if it exists.
The default value of the environmental variable, PATH, is set automatically using SoftEnv. See below for more information.
/bin/tcsh
The file ~/.cshrc is used to customize the user's environment if his login shell is /bin/tcsh.
Softenv
SoftEnv is a utility that is supposed to help users manage complex user environments with potentially conflicting application versions and libraries.
System Default Path
When a user logs in, the system /etc/profile or /etc/csh.cshrc (depending on login shell, and mirrored from csm:/cfmroot/etc/profile) calls /usr/local/packages/softenv-1.6.2/bin/use.softenv.sh to set up the default path via the SoftEnv database.
SoftEnv looks for a user's ~/.soft file and updates the variables and paths accordingly.
Viewing Available Packages
The command softenv will provide a list of available packages. The listing will look something like:
$ softenv These are the macros available: * @default These are the keywords explicitly available: +amber-8 Applications: 'Amber', version: 8 Amber is a +apache-ant-1.6.5 Ant, Java based XML make system version: 1.6. +charm-5.9 Applications: 'Charm++', version: 5.9 Charm++ +default this is the default environment...nukes /etc/ +essl-4.2 Libraries: 'ESSL', version: 4.2 ESSL is a sta +gaussian-03 Applications: 'Gaussian', version: 03 Gaussia ... some stuff deleted ...
Managing SoftEnv
The file ~/.soft in the user's home directory is where the different packages are managed. Add the +keyword into your .soft file. For instance, ff one wants to add the Amber Molecular Dynamics package into their environment, the end of the .soft file should look like this:
+amber-8
@default
To update the environment after modifying this file, one simply uses the resoft command:
% resoft
The command soft can be used to manipulate the environment from the command line. It takes the form:
$ soft add/delete +keyword
Using this method of adding or removing keywords requires the user to pay attention to possible order dependencies. That is, best results require the user to remove keywords in the reverse order in which they were added. It is handy to test out individual keys, but can lead to trouble if changing multiple keys. Changing the .soft file and issuing the resoft is the recommended way of dealing with multiple changes.
▶ Display Module Names for mpip on all clusters.
| Machine | Version | Module |
|---|---|---|
| None Available | N/A | N/A |
▶ Module FAQ?
The information here is applicable to LSU HPC and LONI systems.
Shells
A user may choose between using /bin/bash and /bin/tcsh. Details about each shell follows.
/bin/bash
System resource file: /etc/profile
When one access the shell, the following user files are read in if they exist (in order):
- ~/.bash_profile (anything sent to STDOUT or STDERR will cause things like rsync to break)
- ~/.bashrc (interactive login only)
- ~/.profile
When a user logs out of an interactive session, the file ~/.bash_logout is executed if it exists.
The default value of the environmental variable, PATH, is set automatically using SoftEnv. See below for more information.
/bin/tcsh
The file ~/.cshrc is used to customize the user's environment if his login shell is /bin/tcsh.
Modules
Modules is a utility which helps users manage the complex business of setting up their shell environment in the face of potentially conflicting application versions and libraries.
Default Setup
When a user logs in, the system looks for a file named .modules in their home directory. This file contains module commands to set up the initial shell environment.
Viewing Available Modules
The command
$ module avail
displays a list of all the modules available. The list will look something like:
--- some stuff deleted --- velvet/1.2.10/INTEL-14.0.2 vmatch/2.2.2 ---------------- /usr/local/packages/Modules/modulefiles/admin ----------------- EasyBuild/1.11.1 GCC/4.9.0 INTEL-140-MPICH/3.1.1 EasyBuild/1.13.0 INTEL/14.0.2 INTEL-140-MVAPICH2/2.0 --- some stuff deleted ---
The module names take the form appname/version/compiler, providing the application name, the version, and information about how it was compiled (if needed).
Managing Modules
Besides avail, there are other basic module commands to use for manipulating the environment. These include:
add/load mod1 mod2 ... modn . . . Add modules rm/unload mod1 mod2 ... modn . . Remove modules switch/swap mod . . . . . . . . . Switch or swap one module for another display/show . . . . . . . . . . List modules loaded in the environment avail . . . . . . . . . . . . . . List available module names whatis mod1 mod2 ... modn . . . . Describe listed modules
The -h option to module will list all available commands.
Module is currently available only on SuperMIC.
Usage
Usage involves recompiling a program using the mpiP libraries instead of the standard libraries. Run the program as would be done normally, only look for output files with names of the form:
executable_nm.
▶ Open Example?
@ mpiP @ Command : ./bin/2dheat.x -v -w 100 -h 100 @ Version : 3.1.1 @ MPIP Build date : Oct 16 2007, 09:41:28 @ Start time : 2007 10 16 09:54:22 @ Stop time : 2007 10 16 09:54:37 @ Timer Used : mread_real_time @ MPIP env var : [null] @ Collector Rank : 0 @ Collector PID : 102720 @ Final Output Dir : . @ Report generation : Single collector task @ MPI Task Assignment : 0 l2f1n01.sys.loni.org @ MPI Task Assignment : 1 l2f1n01.sys.loni.org @ MPI Task Assignment : 2 l2f1n01.sys.loni.org @ MPI Task Assignment : 3 l2f1n01.sys.loni.org @ MPI Task Assignment : 4 l2f1n01.sys.loni.org @ MPI Task Assignment : 5 l2f1n01.sys.loni.org @ MPI Task Assignment : 6 l2f1n01.sys.loni.org @ MPI Task Assignment : 7 l2f1n01.sys.loni.org --------------------------------------------------------------------------- @--- MPI Time (seconds) --------------------------------------------------- --------------------------------------------------------------------------- Task AppTime MPITime MPI% 0 15.3 1.02 6.66 1 15.3 0.293 1.91 2 15.3 0.607 3.95 3 15.3 0.239 1.56 4 15.3 0.873 5.69 5 15.3 1.01 6.58 6 15.3 0.646 4.21 7 15.3 1.68 10.94 * 123 6.37 5.19 --------------------------------------------------------------------------- @--- Callsites: 11 -------------------------------------------------------- --------------------------------------------------------------------------- ID Lev File/Address Line Parent_Funct MPI_Call 1 0 0x1000106f [unknown] Reduce 2 0 0x100010d3 [unknown] Bcast 3 0 0x100011f7 [unknown] Barrier 4 0 0x1000242f [unknown] Recv 5 0 0x10002233 [unknown] Isend 6 0 0x100022a7 [unknown] Barrier 7 0 0x10000d1b [unknown] Barrier 8 0 0x1000243b [unknown] Barrier 9 0 0x10000d3b [unknown] Bcast 10 0 0x100023bf [unknown] Isend 11 0 0x1000229b [unknown] Recv --------------------------------------------------------------------------- @--- Aggregate Time (top twenty, descending, milliseconds) ---------------- --------------------------------------------------------------------------- Call Site Time App% MPI% COV Bcast 2 4.18e+03 3.40 65.59 1.03 Reduce 1 1.74e+03 1.42 27.28 1.58 Barrier 3 135 0.11 2.12 0.14 Barrier 6 119 0.10 1.87 0.22 Barrier 8 95.9 0.08 1.51 0.14 Isend 5 32.8 0.03 0.52 0.26 Recv 11 25.9 0.02 0.41 0.65 Recv 4 22.8 0.02 0.36 0.95 Isend 10 21.1 0.02 0.33 0.14 Barrier 7 0.875 0.00 0.01 0.46 Bcast 9 0.591 0.00 0.01 0.30 --------------------------------------------------------------------------- @--- Aggregate Sent Message Size (top twenty, descending, bytes) ---------- --------------------------------------------------------------------------- Call Site Count Total Avrg Sent% Isend 10 5250 2.1e+06 400 49.43 Isend 5 5250 2.1e+06 400 49.43 Reduce 1 6000 2.4e+04 4 0.56 Bcast 2 6000 2.4e+04 4 0.56 Bcast 9 8 32 4 0.00 --------------------------------------------------------------------------- @--- Callsite Time statistics (all, milliseconds): 84 --------------------- --------------------------------------------------------------------------- Name Site Rank Count Max Mean Min App% MPI% Barrier 3 0 749 0.656 0.0203 0.00887 0.10 1.49 Barrier 3 1 749 0.659 0.0226 0.00897 0.11 5.80 Barrier 3 2 749 0.655 0.0204 0.00881 0.10 2.52 Barrier 3 3 749 0.654 0.0222 0.00925 0.11 6.96 Barrier 3 4 749 0.664 0.0267 0.00954 0.13 2.29 Barrier 3 5 749 0.659 0.0265 0.00941 0.13 1.97 Barrier 3 6 749 0.262 0.0236 0.00798 0.12 2.73 Barrier 3 7 749 0.223 0.0179 0.00972 0.09 0.80 Barrier 3 * 5992 0.664 0.0225 0.00798 0.11 2.12 Barrier 6 0 750 0.197 0.0278 0.00677 0.14 2.04 Barrier 6 1 750 0.17 0.0177 0.0066 0.09 4.54 Barrier 6 2 750 0.175 0.0196 0.00648 0.10 2.42 Barrier 6 3 750 0.179 0.0186 0.00663 0.09 5.84 Barrier 6 4 750 0.176 0.0201 0.00707 0.10 1.73 Barrier 6 5 750 0.179 0.0148 0.00665 0.07 1.10 Barrier 6 6 750 0.171 0.0157 0.00673 0.08 1.82 Barrier 6 7 750 0.19 0.0245 0.00827 0.12 1.09 Barrier 6 * 6000 0.197 0.0199 0.00648 0.10 1.87 Barrier 7 0 1 0.0117 0.0117 0.0117 0.00 0.00 Barrier 7 1 1 0.118 0.118 0.118 0.00 0.04 Barrier 7 2 1 0.14 0.14 0.14 0.00 0.02 Barrier 7 3 1 0.11 0.11 0.11 0.00 0.05 Barrier 7 4 1 0.145 0.145 0.145 0.00 0.02 Barrier 7 3 1 0.11 0.11 0.11 0.00 0.05 Barrier 7 4 1 0.145 0.145 0.145 0.00 0.02 Barrier 7 5 1 0.176 0.176 0.176 0.00 0.02 Barrier 7 6 1 0.105 0.105 0.105 0.00 0.02 Barrier 7 7 1 0.0693 0.0693 0.0693 0.00 0.00 Barrier 7 * 8 0.176 0.109 0.0117 0.00 0.01 Barrier 8 0 750 0.137 0.019 0.00846 0.09 1.39 Barrier 8 1 750 0.136 0.0147 0.00692 0.07 3.77 Barrier 8 2 750 0.203 0.0167 0.00771 0.08 2.06 Barrier 8 3 750 0.206 0.0149 0.00692 0.07 4.67 Barrier 8 4 750 0.205 0.0152 0.00686 0.07 1.31 Barrier 8 5 750 0.206 0.0134 0.00675 0.07 1.00 Barrier 8 6 750 0.207 0.0145 0.00682 0.07 1.68 Barrier 8 7 750 0.21 0.0194 0.00649 0.09 0.87 Barrier 8 * 6000 0.21 0.016 0.00649 0.08 1.51 Bcast 2 0 750 0.0523 0.00671 0.00569 0.03 0.49 Bcast 2 1 750 39.8 0.311 0.0146 1.52 79.64 Bcast 2 2 750 39.7 0.169 0.00598 0.82 20.85 Bcast 2 3 750 39.7 0.239 0.0158 1.17 75.12 Bcast 2 4 750 5.54 0.648 0.0114 3.17 55.67 Bcast 2 5 750 6.33 1.27 0.0184 6.19 94.06 Bcast 2 6 750 39.9 0.775 0.0157 3.79 89.90 Bcast 2 7 750 41.6 2.15 0.0318 10.53 96.25 Bcast 2 * 6000 41.6 0.696 0.00569 3.40 65.59 Bcast 9 0 1 0.0374 0.0374 0.0374 0.00 0.00 Bcast 9 1 1 0.0627 0.0627 0.0627 0.00 0.02 Bcast 9 2 1 0.0702 0.0702 0.0702 0.00 0.01 Bcast 9 3 1 0.0948 0.0948 0.0948 0.00 0.04 Bcast 9 4 1 0.0552 0.0552 0.0552 0.00 0.01 Bcast 9 5 1 0.0858 0.0858 0.0858 0.00 0.01 Bcast 9 6 1 0.0782 0.0782 0.0782 0.00 0.01 Bcast 9 7 1 0.107 0.107 0.107 0.00 0.01 Bcast 9 * 8 0.107 0.0739 0.0374 0.00 0.01 Isend 5 0 750 0.153 0.00433 0.00253 0.02 0.32 Isend 5 1 750 0.332 0.00798 0.00271 0.04 2.05 Isend 5 2 750 0.276 0.00507 0.00263 0.02 0.63 Isend 5 3 750 0.313 0.00651 0.00273 0.03 2.05 Isend 5 4 750 0.269 0.00517 0.00263 0.03 0.44 Isend 5 5 750 0.275 0.00871 0.0027 0.04 0.65 Isend 5 6 750 0.272 0.00599 0.00265 0.03 0.69 Isend 5 * 5250 0.332 0.00625 0.00253 0.03 0.52 Isend 10 1 750 0.105 0.00457 0.00242 0.02 1.17 Isend 10 2 750 0.0222 0.00329 0.00236 0.02 0.41 Isend 10 3 750 0.0288 0.00399 0.00244 0.02 1.25 Isend 10 4 750 0.0544 0.00369 0.00242 0.02 0.32 Isend 10 5 750 0.118 0.00487 0.00243 0.02 0.36 Isend 10 6 750 0.0284 0.0036 0.00238 0.02 0.42 Isend 10 7 750 0.128 0.0041 0.00267 0.02 0.18 Isend 10 * 5250 0.128 0.00402 0.00236 0.02 0.33 Recv 4 0 750 0.203 0.0136 0.00259 0.07 1.00 Recv 4 1 750 0.0447 0.00193 0.00154 0.01 0.50 Recv 4 2 750 0.083 0.00278 0.00155 0.01 0.34 Recv 4 3 750 0.0481 0.00241 0.00153 0.01 0.76 Recv 4 4 750 0.12 0.00346 0.00154 0.02 0.30 Recv 4 5 750 0.0709 0.00249 0.00157 0.01 0.18 Recv 4 6 750 0.122 0.00371 0.00158 0.02 0.43 Recv 4 * 5250 0.203 0.00435 0.00153 0.02 0.36 Recv 11 1 750 0.173 0.00274 0.00169 0.01 0.70 Recv 11 2 750 0.183 0.00504 0.00174 0.02 0.62 Recv 11 3 750 0.0227 0.00239 0.00171 0.01 0.75 Recv 11 4 750 0.18 0.00411 0.00173 0.02 0.35 Recv 11 5 750 0.0236 0.00239 0.00174 0.01 0.18 Recv 11 6 750 0.181 0.0066 0.0018 0.03 0.77 Recv 11 7 750 0.275 0.0113 0.00253 0.06 0.50 Recv 11 * 5250 0.275 0.00494 0.00169 0.02 0.41 Reduce 1 0 750 40.8 1.27 0.0178 6.21 93.26 Reduce 1 1 750 0.0167 0.00695 0.0053 0.03 1.78 Reduce 1 2 750 2.56 0.567 0.00809 2.77 70.11 Reduce 1 3 750 0.0217 0.00803 0.00521 0.04 2.52 Reduce 1 4 750 40.5 0.438 0.00794 2.14 37.58 Reduce 1 5 750 0.0202 0.00641 0.00513 0.03 0.48 Reduce 1 6 750 1.35 0.0132 0.0067 0.06 1.53 Reduce 1 7 750 0.0539 0.00646 0.00537 0.03 0.29 Reduce 1 * 6000 40.8 0.29 0.00513 1.42 27.28 --------------------------------------------------------------------------- @--- Callsite Message Sent statistics (all, sent bytes) ------------------- --------------------------------------------------------------------------- Name Site Rank Count Max Mean Min Sum Bcast 2 0 750 4 4 4 3000 Bcast 2 1 750 4 4 4 3000 Bcast 2 2 750 4 4 4 3000 Bcast 2 3 750 4 4 4 3000 Bcast 2 4 750 4 4 4 3000 Bcast 2 5 750 4 4 4 3000 Bcast 2 6 750 4 4 4 3000 Bcast 2 7 750 4 4 4 3000 Bcast 2 * 6000 4 4 4 2.4e+04 Bcast 9 0 1 4 4 4 4 Bcast 9 1 1 4 4 4 4 Bcast 9 2 1 4 4 4 4 Bcast 9 3 1 4 4 4 4 Bcast 9 4 1 4 4 4 4 Bcast 9 5 1 4 4 4 4 Bcast 2 6 750 4 4 4 3000 Bcast 2 7 750 4 4 4 3000 Bcast 2 * 6000 4 4 4 2.4e+04 Bcast 9 0 1 4 4 4 4 Bcast 9 1 1 4 4 4 4 Bcast 9 2 1 4 4 4 4 Bcast 9 3 1 4 4 4 4 Bcast 9 4 1 4 4 4 4 Bcast 9 5 1 4 4 4 4 Bcast 9 6 1 4 4 4 4 Bcast 9 7 1 4 4 4 4 Bcast 9 * 8 4 4 4 32 Isend 5 0 750 400 400 400 3e+05 Isend 5 1 750 400 400 400 3e+05 Isend 5 2 750 400 400 400 3e+05 Isend 5 3 750 400 400 400 3e+05 Isend 5 4 750 400 400 400 3e+05 Isend 5 5 750 400 400 400 3e+05 Isend 5 6 750 400 400 400 3e+05 Isend 5 * 5250 400 400 400 2.1e+06 Isend 10 1 750 400 400 400 3e+05 Isend 10 2 750 400 400 400 3e+05 Isend 10 3 750 400 400 400 3e+05 Isend 10 4 750 400 400 400 3e+05 Isend 10 5 750 400 400 400 3e+05 Isend 10 6 750 400 400 400 3e+05 Isend 10 7 750 400 400 400 3e+05 Isend 10 * 5250 400 400 400 2.1e+06 Reduce 1 0 750 4 4 4 3000 Reduce 1 1 750 4 4 4 3000 Reduce 1 2 750 4 4 4 3000 Reduce 1 3 750 4 4 4 3000 Reduce 1 4 750 4 4 4 3000 Reduce 1 5 750 4 4 4 3000 Reduce 1 6 750 4 4 4 3000 Reduce 1 7 750 4 4 4 3000 Reduce 1 * 6000 4 4 4 2.4e+04 --------------------------------------------------------------------------- @--- End of Report -------------------------------------------------------- ---------------------------------------------------------------------------
Resources
Last modified: August 21 2017 10:47:37.