1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
|
<!doctype linuxdoc system>
<article>
<title>Apache Web Server Performance Tuning
<abstract>
In this window, you can set the Web server to optimize trade-offs
between performance and CPU usage. Normally, however, the default
values are good enough.
</abstract>
<sect>Number of Processes
<p>
Apache is built around a parent process, which accepts connections from
clients, and one or more child processes. These child processes
complete the requests and return the accessed page. The number of child
processes changes dynamically, according to the number of connections to
the Web server. The number of processes is limited by the value
specified here.
<sect1>Start servers
<p>
The Web server will start with the number of child processes
specified here.
<sect1>Max clients per servers
<p>
The maximum number of child processes. Each client connection uses a
child process, so this value indirectly controls how many clients can
simultaneously connect to the Web server. If this value is too low,
clients will be locked out, so be generous. This value is intended
mostly as a safeguard against having a runaway web server crash the
entire system.
<sect1>Max requests per child
<p>
A child process will exit after serving this number of
accesses. It is intended as a failsafe against possible
memory leaks. Currently, however, there are none known.
<sect1>Minimum of spare servers
<p>
Apache makes sure that there always are a few idle child processes, for
handling transient load spikes. If fewer than this number of child
processes are idle, more are started.
<sect1>Maximum of spare servers
<p>
Apache also makes sure that no more than this number of idle child
processes exist. If more exist, some of them are killed.
<sect>Keeps Alive
<p>
To reduce the bandwidth used by Web traffic, HTTP was extended to allow
more than one request within a single connection. This is called a
<em/persistent connection/ or you can say that the connection is
<em/kept alive/.
<sect1>Keep Alive
<p>
If enabled, browsers can establish persistent connections. If disabled,
the Web server will accept only one access per connection.
<sect1>Keep Alive Timeout
<p>
This specifies the number of seconds to wait for the next request
in a persistent connection. If it is exceeded, the connection is
closed.
<sect1>Max Keep Alive Requests
<p>
The maximum number of requests allowed during a persistent connection.
<sect>Advanced Tuning
<p>
<sect1>Time out
<p>
The number of seconds the Web server will wait for a request to be sent,
after a client has connected to the server. If a client connects, but
does not send any requests, the client will be disconnected after this
time (freeing the resources it was occupying).
</article>
|
