From cc9452e346694f69288ba8d152f41fa65e475ff8 Mon Sep 17 00:00:00 2001 From: Frank Baker Date: Thu, 2 Aug 2001 11:10:11 -0500 Subject: [svn-r4303] Purpose: Bringing 'last minute' release branch edits into development branch. Description: Chunking.html, Filters.html, and Performance.html Added links, with explanatory note, from the first 2 files above to the new freespace management section in Performance.html. Platforms tested: IE 5 --- doc/html/Chunking.html | 22 +++++++++++++++++++--- doc/html/Filters.html | 20 ++++++++++++++++++-- doc/html/Performance.html | 4 +++- 3 files changed, 40 insertions(+), 6 deletions(-) diff --git a/doc/html/Chunking.html b/doc/html/Chunking.html index f898f58..8fd363a 100644 --- a/doc/html/Chunking.html +++ b/doc/html/Chunking.html @@ -58,6 +58,7 @@
  • 3. Cache Efficiency
  • 4. Fragmentation
  • 5. File Storage Overhead +
  • 6. Chunk Compression

    1. Introduction

    @@ -215,7 +216,22 @@ H5Pset_istore_k() (increasing the out degree actually increases file overhead while decreasing the number of nodes). -
    + +

    6. Chunk Compression

    + +

    Dataset chunks can be compressed through the use of filters. + See the chapter “Filters in HDF5.” + +

    Reading and rewriting compressed chunked data can result in holes + in an HDF5 file. In time, enough such holes can increase the + file size enough to impair application or library performance + when working with that file. See + “Freespace Management” + in the chapter + “Performance Analysis and Issues.” + + +

    Footnote 1: Parallel versions of the library can access individual bytes of a chunk when the underlying file @@ -224,7 +240,6 @@

    Footnote 2: The raw data chunk cache was added before the second alpha release. -

    @@ -273,12 +288,13 @@ -Last modified: 16 February 2001 +Last modified: 2 August 2001
    Describes HDF5 Release 1.5, Unreleased Development Branch + diff --git a/doc/html/Filters.html b/doc/html/Filters.html index e4efc5f..f292bfd 100644 --- a/doc/html/Filters.html +++ b/doc/html/Filters.html @@ -55,7 +55,8 @@

    1. Introduction

    -

    HDF5 allows chunked data to pass through user-defined filters +

    HDF5 allows chunked data1 + to pass through user-defined filters on the way to or from disk. The filters operate on chunks of an H5D_CHUNKED dataset can be arranged in a pipeline so output of one filter becomes the input of the next filter. @@ -507,6 +508,21 @@ H5Z: filter statistics accumulated over life of library:

    + + +

    Footnote 1: Dataset chunks can be compressed + through the use of filters. Developers should be aware that + reading and rewriting compressed chunked data can result in holes + in an HDF5 file. In time, enough such holes can increase the + file size enough to impair application or library performance + when working with that file. See + “Freespace Management” + in the chapter + “Performance Analysis and Issues.” + + + +

    @@ -554,7 +570,7 @@ H5Z: filter statistics accumulated over life of library: -Last modified: 17 December 1999 +Last modified: 2 August 2001
    diff --git a/doc/html/Performance.html b/doc/html/Performance.html index 36accbf..5486e57 100644 --- a/doc/html/Performance.html +++ b/doc/html/Performance.html @@ -63,7 +63,9 @@ Dataset Chunking Issues elsewhere in this User's Guide. +

    3. Freespace Management

    +

    HDF5 does not yet manage freespace as effectively as it might. While a file is opened, the library actively tracks and re-uses @@ -235,7 +237,7 @@ -Last modified: 11 July 2001 +Last modified: 2 August 2001
    -- cgit v0.12