iTP Secure WebServer System Administrator's Guide (Version 7.0)

Configuration Directives
iTP Secure WebServer System Administrator’s Guide—523346-012
A-27
Default
Persistent connections are a feature introduced in the HTTP/1.1 protocol to improve
performance. In earlier versions of HTTP, each request for a URL resulted in a new
connection. Serving a page often requires many requests (for example to include a
graphics file in a page of text), so a complex page could take considerable time to load.
When a server supports persistent connections, it establishes a connection when the
user makes a request; the connection stays open for the series of related requests the
client makes of the server.
This directive enables you to control how long the server waits for the next request
from the client. If the timeout expires, the server closes the connection. If a new
request arrives from the client, the server creates a new connection. The user does not
experience any disruption of service.
The timeout-value has a range from 0 seconds to the value of the InputTimeout
directive. A value of 0 causes the server not to create persistent connections; that is,
the server will behave as in previous releases.
Default
KeepAliveTimeout 15
Example
KeepAliveTimeout 30
KeepAliveMaxRequest
Syntax
KeepAliveMaxRequest integer-value
Description
Use the KeepAliveMaxRequest directive to specify the number of requests the iTP
Secure WebServer should handle before closing a persistent connection.
Persistent connections are a feature introduced in the HTTP/1.1 protocol to improve
performance. In earlier versions of HTTP, each request for a URL resulted in a new
connection. Serving a page often requires many requests (for example to include a
graphics file in a page of text), so a complex page could take considerable time to load.
When a server supports persistent connections, it establishes a connection when the
user makes a request; the connection stays open for the series of related requests the
client makes of the server.
This directive enables you to control how many requests the server will accept on the
same connection. If the number is exceeded, the server closes the connection; when
the next request arrives, the server creates a new connection. The user does not
experience any disruption of service.