Merge branch 'develop' into logging_typo_vpicio_mts

Author: jeanbez

.gitlab-ci.yml

@@ -457,7 +457,7 @@ perlmutter-metrics:
         PDC_TMPDIR: "${CI_BUILDS_DIR}/${CI_PROJECT_NAME}/${CI_JOB_ID}/pdc-tmp-metrics"
         PDC_DATA_LOC: "${CI_BUILDS_DIR}/${CI_PROJECT_NAME}/${CI_JOB_ID}/pdc-data-metrics"
         PDC_CLIENT_LOOKUP: "NONE"
-        PDC_SERVER: "${PDC_BUILD_PATH}/perlmutter/metrics/bin/pdc_server.exe"
+        PDC_SERVER: "${PDC_BUILD_PATH}/perlmutter/metrics/bin/pdc_server"
         PDC_SERVER_CLOSE: "${PDC_BUILD_PATH}/perlmutter/metrics/bin/close_server"
         PDC_CLIENT: "${PDC_BUILD_PATH}/perlmutter/metrics/bin/vpicio_mts"
         PDC_JOB_OUTPUT: "pdc-metrics.log"

CMakeLists.txt

@@ -637,7 +637,7 @@ install(
 
 #install(
 #  FILES
-#    ${PDC_BINARY_DIR}/bin/pdc_server.exe
+#    ${PDC_BINARY_DIR}/bin/pdc_server
 #    ${PDC_BINARY_DIR}/bin/close_server
 #  DESTINATION
 #    ${CMAKE_INSTALL_PREFIX}/bin

docs/source/developer-notes.rst

@@ -595,7 +595,7 @@ To run a test, let's lunch pdc_server with 4 cores:
 
     cd $PDC_HOME/build
     rm -rf ./pdc_tmp # optional if you need to clean up the PDC tmp directory
-    srun -N 1 -n  4 -c 2 --mem=25600 --cpu_bind=cores ./bin/pdc_server.exe &
+    srun -N 1 -n  4 -c 2 --mem=25600 --cpu_bind=cores ./bin/pdc_server &
 
 To debug the client, you can run the following command:
 
@@ -610,6 +610,6 @@ But if you need to debug the server, you can prepend ``srun`` with ``ddt --conne
 
     cd $PDC_HOME/build
     rm -rf ./pdc_tmp # optional if you need to clean up the PDC tmp directory
-    ddt --connect srun -N 1 -n  4 -c 2 --mem=25600 --cpu_bind=cores ./bin/pdc_server.exe &
+    ddt --connect srun -N 1 -n  4 -c 2 --mem=25600 --cpu_bind=cores ./bin/pdc_server &
 
 We recommend to use 1 node when debugging PDC, but if memory is not sufficient, you can use more nodes. 

docs/source/getting_started.rst

@@ -276,7 +276,7 @@ Run 2 server processes in the background
 
 .. code-block:: Bash
 
-	mpiexec -np 2 $PDC_DIR/bin/pdc_server.exe &
+	mpiexec -np 2 $PDC_DIR/bin/pdc_server &
 
 Run 4 client processes that concurrently create 1000 objects and then create and query 1000 tags:
 
@@ -291,7 +291,7 @@ Run 4 server processes, each on one compute node in the background:
 
 .. code-block:: Bash
 
-	srun -N 4 -n 4 -c 2 --mem=25600 --cpu_bind=cores $PDC_DIR/bin/pdc_server.exe &
+	srun -N 4 -n 4 -c 2 --mem=25600 --cpu_bind=cores $PDC_DIR/bin/pdc_server &
 
 Run 64 client processes that concurrently create 1000 objects and then create and query 100000 tags:
 

docs/source/hdf5vol.rst

@@ -45,7 +45,7 @@ To compile and run examples:
 	$ cd vol-pdc/examples
 	$ cmake .
 	$ make
-	$ mpirun -N 1 -n 1 -c 1 <pdc installation directory>/bin/pdc_server.exe &
+	$ mpirun -N 1 -n 1 -c 1 <pdc installation directory>/bin/pdc_server &
 	$ mpirun -N 1 -n 1 -c 1 ./h5pdc_vpicio test
 	$ mpirun -N 1 -n 1 -c 1 <pdc installation directory>/bin/close_server
 

docs/source/tools.rst

@@ -163,7 +163,7 @@ Examples:
 
 .. code-block:: Bash
 
-	$ srun -N 1 -n 1 -c 2 --mem=25600 --cpu_bind=cores --gres=craynetwork:1 --overlap /path/to/pdc_server.exe &
+	$ srun -N 1 -n 1 -c 2 --mem=25600 --cpu_bind=cores --gres=craynetwork:1 --overlap /path/to/pdc_server &
 	==PDC_SERVER[0]: using [./pdc_tmp/] as tmp dir, 1 OSTs, 1 OSTs per data file, 0% to BB
 	==PDC_SERVER[0]: using ofi+tcp
 	==PDC_SERVER[0]: without multi-thread!
@@ -201,7 +201,7 @@ Examples:
 
 .. code-block:: Bash
 
-	$ srun -N 1 -n 1 -c 2 --mem=25600 --cpu_bind=cores --gres=craynetwork:1 --overlap /path/to/pdc_server.exe &
+	$ srun -N 1 -n 1 -c 2 --mem=25600 --cpu_bind=cores --gres=craynetwork:1 --overlap /path/to/pdc_server &
 	==PDC_SERVER[0]: using [./pdc_tmp/] as tmp dir, 1 OSTs, 1 OSTs per data file, 0% to BB
 	==PDC_SERVER[0]: using ofi+tcp
 	==PDC_SERVER[0]: without multi-thread!

examples/README.md

@@ -81,6 +81,6 @@
   + This multidataset_plugin implementation is designed for wrapping HDF5 operations and call HDF5 multidataset. We demonstrate that PDC can substitute a storage structure comparable to HDF5. Therefore, it is possible to make direct comparison between these two. Currently, only write is implemented in this example.
 ```
 0. cd C_plus_plus_example;make
-1. mpiexec -n 1 ./pdc_server.exe &
+1. mpiexec -n 1 ./pdc_server &
 2. mpiexec -n 1 ./region_transfer_1D_append
 ```

examples/llsm/LLSM_IMPORTER.md

@@ -58,7 +58,7 @@ cd $WORK_SPACE/source/pdc/scripts/llsm_importer
 
 Modify the template script `template.sh`. 
 
-Change `EXECPATH` to where your `pdc_server.exe` is installed
+Change `EXECPATH` to where your `pdc_server` is installed
 Change `TOOLPATH` to where your `llsm_importer` artifact is.
 
 Change `LLSM_DATA_PATH` to where your sample dataset is. For exmaple, 

examples/mpi_test.sh

@@ -22,8 +22,8 @@ echo $test_args
 if [ -x $test_exe ]; then echo "testing: $test_exe"; else echo "test: $test_exe not found or not and executable" && exit -2; fi
 rm -rf pdc_tmp pdc_data
 # START the server (in the background)
-echo "$mpi_cmd -n $n_servers $extra_cmd ./pdc_server.exe &"
-$mpi_cmd -n $n_servers $extra_cmd ./pdc_server.exe &
+echo "$mpi_cmd -n $n_servers $extra_cmd ./pdc_server &"
+$mpi_cmd -n $n_servers $extra_cmd ./pdc_server &
 # WAIT a bit, for 1 second
 sleep 1
 # RUN the actual test

examples/multiple_mpi_test.sh

@@ -22,8 +22,8 @@ n_client="$5"
 test_args="$6 $7 $8"
 # if [ -x $test_exe ]; then echo "testing: $test_exe"; else echo "test: $test_exe not found or not and executable" && exit -2; fi
 # START the server (in the background)
-echo "$mpi_cmd -n $n_servers $extra_cmd ./pdc_server.exe &"
-$mpi_cmd -n $n_servers $extra_cmd $PDC_DIR/bin/./pdc_server.exe &
+echo "$mpi_cmd -n $n_servers $extra_cmd ./pdc_server &"
+$mpi_cmd -n $n_servers $extra_cmd $PDC_DIR/bin/./pdc_server &
 # WAIT a bit...
 sleep 1
 # RUN the actual test(s)