[Darshan-commits] [Git][darshan/darshan][dev-detailed-hdf5-mod] 2 commits: fix version warnings for H5F module

Shane Snyder xgitlab at cels.anl.gov
Fri Apr 3 10:54:43 CDT 2020 


Shane Snyder pushed to branch dev-detailed-hdf5-mod at darshan / darshan


Commits:
66b9c997 by Shane Snyder at 2020-04-02T13:40:08-05:00
fix version warnings for H5F module

- - - - -
8e0847f7 by Shane Snyder at 2020-04-03T10:54:24-05:00
updated HDF5 documentation for darshan-util

- - - - -


2 changed files:

- darshan-util/darshan-hdf5-logutils.c
- darshan-util/doc/darshan-util.txt


Changes:

=====================================
darshan-util/darshan-hdf5-logutils.c
=====================================
@@ -370,6 +370,8 @@ static void darshan_log_print_hdf5_file_description(int ver)
     if(ver <= 2)
     {
         printf("\n# WARNING: H5F module log format version <=2 does not support the following counters:\n");
+        printf("# \t- H5F_FLUSHES\n");
+        printf("# \t- H5F_USE_MPIIO\n");
         printf("# \t- H5F_F_META_TIME\n");
     }
 


=====================================
darshan-util/doc/darshan-util.txt
=====================================
@@ -321,9 +321,9 @@ value of 1 MiB for optimal file alignment.
 | MPIIO_SLOWEST_RANK_BYTES | The number of bytes transferred by the rank with the largest time spent in MPI I/O
 | MPIIO_F_*_START_TIMESTAMP | Timestamp that the first MPIIO file open/read/write/close operation began
 | MPIIO_F_*_END_TIMESTAMP | Timestamp that the last MPIIO file open/read/write/close operation ended
-| MPIIO_READ_TIME | Cumulative time spent reading at MPI level
-| MPIIO_WRITE_TIME | Cumulative time spent write and sync at MPI level
-| MPIIO_META_TIME | Cumulative time spent in open and close at MPI level
+| MPIIO_F_READ_TIME | Cumulative time spent reading at MPI level
+| MPIIO_F_WRITE_TIME | Cumulative time spent write and sync at MPI level
+| MPIIO_F_META_TIME | Cumulative time spent in open and close at MPI level
 | MPIIO_F_MAX_READ_TIME | Duration of the slowest individual MPI read operation
 | MPIIO_F_MAX_WRITE_TIME | Duration of the slowest individual MPI write operation
 | MPIIO_F_FASTEST_RANK_TIME | The time of the rank which had the smallest amount of time spent in MPI I/O (cumulative read, write, and meta times)
@@ -362,15 +362,64 @@ value of 1 MiB for optimal file alignment.
 | STDIO_F_VARIANCE_RANK_BYTES | The population variance for bytes transferred of all the ranks
 |====
 
-.HDF5 module
+.H5F module
 [cols="40%,60%",options="header"]
 |====
 | counter name | description
-| HDF5_OPENS | Count of HDF5 opens
-| HDF5_F_*_START_TIMESTAMP | Timestamp that the first HDF5 file open/close operation began
-| HDF5_F_*_END_TIMESTAMP | Timestamp that the last HDF5 file open/close operation ended
+| H5F_OPENS | Count of H5F opens
+| H5F_FLUSHES | Count of H5F flushes
+| H5F_USE_MPIIO | Flag indicating whether MPI-IO is used for accessing the file
+| H5F_F_*_START_TIMESTAMP | Timestamp that the first H5F open/close operation began
+| H5F_F_*_END_TIMESTAMP | Timestamp that the last H5F open/close operation ended
+| H5F_F_META_TIME | Cumulative time spent in H5F open/close/flush operations
 |====
 
+.H5D module
+[cols="40%,60%",options="header"]
+|====
+| counter name | description
+| H5D_OPENS | Count of H5D opens
+| H5D_READS | Count of H5D reads
+| H5D_WRITES | Count of H5D writes
+| H5D_FLUSHES | Count of H5D flushes
+| H5D_BYTES_READ | Total number of bytes read from the dataset using H5D
+| H5D_BYTES_WRITTEN | Total number of bytes written to the dataset using H5D
+| H5D_RW_SWITCHES | Number of times that access toggled between read and write in consecutive H5D operations
+| H5D_REGULAR_HYPERSLAB_SELECTS | Number of H5D read/write ops with regular hyperslab selections
+| H5D_IRREGULAR_HYPERSLAB_SELECTS | Number of H5D read/write ops with irregular hyperslab selections
+| H5D_POINT_SELECTS | Number of read/write ops with point selections
+| H5D_MAX_READ_TIME_SIZE | Size of the slowest H5D read operation
+| H5D_MAX_WRITE_TIME_SIZE | Size of the slowest H5D write operation
+| H5D_SIZE_READ_AGG_* | Histogram of total size of read accesses at H5D level
+| H5D_SIZE_WRITE_AGG_* | Histogram of total size of write accesses at H5D level
+| H5D_ACCESS[1-4]_ACCESS | Sizes of 4 most common H5D accesses
+| H5D_ACCESS[1-4]_LENGTH_D[1-5] | Access lengths along last 5 dimensions (D5 is fastest changing) of 4 most common H5D accesses
+| H5D_ACCESS[1-4]_STRIDE_D[1-5] | Access strides along last 5 dimensions (D5 is fastest changing) of 4 most common H5D accesses
+| H5D_ACCESS[1-4]_COUNT | Count of 4 most common H5D aggregate access sizes
+| H5D_DATASPACE_NDIMS | Number of dimensions in dataset's dataspace
+| H5D_DATASPACE_NPOINTS | Number of points in dataset's dataspace
+| H5D_DATATYPE_SIZE | Total size of dataset elements in bytes
+| H5D_CHUNK_SIZE_D[1-5] | Chunk sizes in the last 5 dimensions of the dataset (D5 is the fastest changing dimension)
+| H5D_USE_MPIIO_COLLECTIVE | Flag indicating use of MPI-IO collectives
+| H5D_USE_DEPRECATED | Flag indicating whether deprecated create/open calls were used
+| H5D_FASTEST_RANK | The MPI rank with smallest time spent in H5D I/O
+| H5D_FASTEST_RANK_BYTES | The number of bytes transferred by the rank with smallest time spent in H5D I/O
+| H5D_SLOWEST_RANK | The MPI rank with largest time spent in H5D I/O
+| H5D_SLOWEST_RANK_BYTES | The number of bytes transferred by the rank with the largest time spent in H5D I/O
+| H5D_F_*_START_TIMESTAMP | Timestamp that the first H5D open/read/write/close operation began
+| H5D_F_*_END_TIMESTAMP | Timestamp that the last H5D open/read/write/close operation ended
+| H5D_F_READ_TIME | Cumulative time spent reading at H5D level
+| H5D_F_WRITE_TIME | Cumulative time spent writing at H5D level
+| H5D_F_META_TIME | Cumulative time spent in open/close/flush at H5D level
+| H5D_F_MAX_READ_TIME | Duration of the slowest individual H5D read operation
+| H5D_F_MAX_WRITE_TIME | Duration of the slowest individual H5D write operation
+| H5D_F_FASTEST_RANK_TIME | The time of the rank which had the smallest amount of time spent in H5D I/O (cumulative read, write, and meta times)
+| H5D_F_SLOWEST_RANK_TIME | The time of the rank which had the largest amount of time spent in H5D I/O
+| H5D_F_VARIANCE_RANK_TIME | The population variance for H5D I/O time of all the ranks
+| H5D_F_VARIANCE_RANK_BYTES | The population variance for bytes transferred of all the ranks at H5D level
+|====
+
+.PnetCDF module
 .PnetCDF module
 [cols="40%,60%",options="header"]
 |====



View it on GitLab: https://xgitlab.cels.anl.gov/darshan/darshan/compare/a3e75fe8b657da3ffdb72abef3a8dadf95da8145...8e0847f7006225ff6f700156ff62cfbb066e5e0c

-- 
View it on GitLab: https://xgitlab.cels.anl.gov/darshan/darshan/compare/a3e75fe8b657da3ffdb72abef3a8dadf95da8145...8e0847f7006225ff6f700156ff62cfbb066e5e0c
You're receiving this email because of your account on xgitlab.cels.anl.gov.


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/darshan-commits/attachments/20200403/eb021eb0/attachment-0001.html>

More information about the Darshan-commits mailing list