Documentation Source Text

Check-in [7e41434c2b]
Login

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

Overview
Comment:Update the 35%-faster document to the latest from trunk.
Timelines: family | ancestors | branch-3.18
Files: files | file ages | folders
SHA3-256: 7e41434c2b25dbfed5e47f84cfbb00f503e080365be344d39742f93c516073c9
User & Date: drh 2017-05-10 16:38:53
Context
2017-05-10
16:38
Update the 35%-faster document to the latest from trunk. Leaf check-in: 7e41434c2b user: drh tags: branch-3.18
2017-05-05
16:56
Copy the 35% faster changes from trunk. check-in: 6fa9ca7f89 user: drh tags: branch-3.18
Changes
Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to pages/fasterthanfs.in.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
<title>35% Faster Than The Filesystem</title>
<tcl>hd_keywords {faster than the filesystem}</tcl>

<table_of_contents>

<h1>Summary</h1>

<p>10 kilobyte blobs (for example, thumbnail
images) can be read out of an SQLite database about 35% faster
than they can be read from individual files on disk.

<p>Furthermore, a single SQLite database holding many thousands
of thumbnail-sized blobs uses about 20% less disk space than
storing the blobs in individual files.

<p>The performance difference arises (we believe) because when
reading from an SQLite database, the open() and close() system calls
are invoked only once, whereas
open() and close() are invoked once for each blob
when reading the blobs from individual files.  It appears that the







|
|


|
|







1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
<title>35% Faster Than The Filesystem</title>
<tcl>hd_keywords {faster than the filesystem}</tcl>

<table_of_contents>

<h1>Summary</h1>

<p>Small blobs (for example, thumbnail images)
can be read out of an SQLite database about 35% faster
than they can be read from individual files on disk.

<p>Furthermore, a single SQLite database holding
10-kilobyte blobs uses about 20% less disk space than
storing the blobs in individual files.

<p>The performance difference arises (we believe) because when
reading from an SQLite database, the open() and close() system calls
are invoked only once, whereas
open() and close() are invoked once for each blob
when reading the blobs from individual files.  It appears that the