added write throttling for large concurency jobs with MPI-POSIX VFD

.gitattributes

@@ -10,6 +10,7 @@
 /README -text
 /autogen.sh -text
 /config.guess -text
+/config.h.in -text
 /config.sub -text
 /configure.ac -text
 /depcomp -text

config.h.in

@@ -0,0 +1,99 @@
+/* config.h.in.  Generated from configure.ac by autoheader.  */
+
+/* Define to 1 if you have the <inttypes.h> header file. */
+#undef HAVE_INTTYPES_H
+
+/* Define to 1 if you have the <memory.h> header file. */
+#undef HAVE_MEMORY_H
+
+/* Define to 1 if you have the <stdint.h> header file. */
+#undef HAVE_STDINT_H
+
+/* Define to 1 if you have the <stdlib.h> header file. */
+#undef HAVE_STDLIB_H
+
+/* Define to 1 if you have the <strings.h> header file. */
+#undef HAVE_STRINGS_H
+
+/* Define to 1 if you have the <string.h> header file. */
+#undef HAVE_STRING_H
+
+/* Define to 1 if you have the <sys/stat.h> header file. */
+#undef HAVE_SYS_STAT_H
+
+/* Define to 1 if you have the <sys/types.h> header file. */
+#undef HAVE_SYS_TYPES_H
+
+/* Define to 1 if you have the <unistd.h> header file. */
+#undef HAVE_UNISTD_H
+
+/* "" */
+#undef MY_BUILD_CPU
+
+/* "" */
+#undef MY_BUILD_OS
+
+/* "" */
+#undef MY_BUILD_VENDOR
+
+/* "" */
+#undef MY_GNUNAME
+
+/* "" */
+#undef MY_UNAME
+
+/* Name of package */
+#undef PACKAGE
+
+/* Define to the address where bug reports for this package should be sent. */
+#undef PACKAGE_BUGREPORT
+
+/* Define to the full name of this package. */
+#undef PACKAGE_NAME
+
+/* Define to the full name and version of this package. */
+#undef PACKAGE_STRING
+
+/* Define to the one symbol short name of this package. */
+#undef PACKAGE_TARNAME
+
+/* Define to the version of this package. */
+#undef PACKAGE_VERSION
+
+/* Define to 1 if you have the ANSI C header files. */
+#undef STDC_HEADERS
+
+/* Enable extensions on AIX 3, Interix.  */
+#ifndef _ALL_SOURCE
+# undef _ALL_SOURCE
+#endif
+/* Enable GNU extensions on systems that have them.  */
+#ifndef _GNU_SOURCE
+# undef _GNU_SOURCE
+#endif
+/* Enable threading extensions on Solaris.  */
+#ifndef _POSIX_PTHREAD_SEMANTICS
+# undef _POSIX_PTHREAD_SEMANTICS
+#endif
+/* Enable extensions on HP NonStop.  */
+#ifndef _TANDEM_SOURCE
+# undef _TANDEM_SOURCE
+#endif
+/* Enable general extensions on Solaris.  */
+#ifndef __EXTENSIONS__
+# undef __EXTENSIONS__
+#endif
+
+
+/* Version number of package */
+#undef VERSION
+
+/* Define to 1 if on MINIX. */
+#undef _MINIX
+
+/* Define to 2 if the system does not provide POSIX.1 features except with
+   this defined. */
+#undef _POSIX_1_SOURCE
+
+/* Define to 1 if you need to in order for `stat' and other things to work. */
+#undef _POSIX_SOURCE

src/H5MultiBlock.c

@@ -998,6 +998,30 @@ _H5MultiBlock_write_data (
 #endif
 	if ( dataset < 0 ) return HANDLE_H5D_CREATE_ERR ( name, f->timestep );
 
+#if H5PART_THROTTLE_WRITES > 0
+	int ret;
+	int token = 1;
+	if (f->myproc == 0) {
+		_H5Part_print_info ("Using throttled writes with factor = %d",
+			H5PART_THROTTLE_WRITES);
+	}
+	if (f->myproc % H5PART_THROTTLE_WRITES > 0) {
+		_H5Part_print_debug_detail (
+			"[%d] Waiting on write token from %d",
+			f->myproc, f->myproc - 1);
+		// wait to receive token before continuing with write
+		ret = MPI_Recv(
+			&token, 1, MPI_INT,
+			f->myproc - 1, // receive from previous proc
+			f->myproc, // use this proc id as message tag
+			f->comm,
+			MPI_STATUS_IGNORE
+			);
+		if ( ret != MPI_SUCCESS ) return HANDLE_MPI_SENDRECV_ERR;
+	}
+	_H5Part_print_debug_detail ("[%d] Executing write", f->myproc);
+#endif
+		
 	herr = H5Dwrite ( 
 		dataset,
 		type,
@@ -1007,6 +1031,24 @@ _H5MultiBlock_write_data (
 		data );
 	if ( herr < 0 ) return HANDLE_H5D_WRITE_ERR ( name, f->timestep );
 
+#if H5PART_THROTTLE_WRITES > 0
+	if (f->myproc % H5PART_THROTTLE_WRITES < H5PART_THROTTLE_WRITES - 1) {
+		// pass token to next proc 
+		if (f->myproc + 1 < f->nprocs) {
+			_H5Part_print_debug_detail (
+				"[%d] Passing write token to %d",
+				f->myproc, f->myproc + 1);
+			ret = MPI_Send(
+				&token, 1, MPI_INT,
+				f->myproc + 1, // send to next proc
+				f->myproc + 1, // use the id of the target as tag
+				f->comm
+				);
+		}
+		if ( ret != MPI_SUCCESS ) return HANDLE_MPI_SENDRECV_ERR;
+	}
+#endif
+
 	herr = H5Dclose ( dataset );
 	if ( herr < 0 ) return HANDLE_H5D_CLOSE_ERR;
 

src/H5Part.c

@@ -1687,7 +1687,7 @@ _H5Part_set_step (
 	f->timestep = step;
 
 	if( f->mode & H5PART_READ ) {
-		_H5Part_print_info (
+		_H5Part_print_debug (
 			"Proc[%d]: Set step to #%lld for file %lld",
 			f->myproc,
 			(long long)step,