Next: NCSA server directives Up: General configuration directives Previous: General configuration directives

CERN server directives

Directives specifying a period of time can have the period entered in any combination of units: seconds, minutes, hours, days, weeks, months (30 days), years (365 days) or for ever (10 years)! These can be abbreviated to sec, min, hour, day, week, month, year or forever. Any characters after these abbreviated forms are ignored.

HostName host-name                (determined automatically)

The host-name of the computer system on which the server is running. This directive is necessary if the server cannot determine the host-name from the system, or if the server system is known by an alias rather than the system's host-name.

    HostName    www.organization.domain

Equivalent to the NCSA ServerName directive.

Port port-no                                    (no default)

The port number on which the server should listen for requests. Note that port numbers less than 1024 are reserved ports: usually processes must be run as root to bind to these ports. Network programs started by non-privileged users can usually use non-reserved ports above 1023. If no port is specified, the server assumes that it has been started from inetd, reads from the standard input stream and outputs to the standard output stream. This feature can be useful for debugging.

    Port            8080

Equivalent to the NCSA Port directive.

UserId user-id                                      (nobody)

The user name or number to which the server should change before accessing files if it was started as root.

    UserId WWW

Equivalent to the NCSA User directive.

GroupId group-id                                   (nogroup)

The group name or number to which the server should change before accessing files if it was started as root.

    GroupId 62534

Equivalent to the NCSA Group directive.

PidFile filename                            (/tmp/httpd-pid)

The path of the file in which the server stores its process ID. This information is used by the server program when the -restart option is given to restart another server that is already running.

    PidFile /tmp/cern_httpd.pid

Equivalent to the NCSA PidFile directive.

Enable method                           (GET, HEAD and POST)

Specifies an HTTP method that should be enabled. The directive can occur more than once.

    Enable DELETE

There is no equivalent NCSA directive.

Disable method               (all except GET, HEAD and POST)

Specifies an HTTP method that should be disabled. The directive can occur more than once.

    Disable POST

There is no equivalent NCSA directive.

IdentityCheck { on  | off }                            (off)

If enabled the server will contact the ident daemon on the system from which the request originated to determine the user name associated with the process issuing the request. This information is then included in log messages. The information cannot be relied upon as being valid: many systems do not run an ident daemon and obtaining the information consumes resources, so it is generally recommended that this option be left disabled.
Equivalent to the CERN IdentityCheck directive.

InputTimeout time-spec                           (2 minutes)

The time allowed for the client to send the HTTP request after the connection has been established, before the server will drop the connection.

    InputTimeout 2 mins 30 secs

Equivalent to the NCSA Timeout directive.

OutputTimeout time-spec                         (20 minutes)

The time allowed for output to be sent to the client.

    OutputTimeout 1 hour

Equivalent to the NCSA Timeout directive.

ScriptTimeout time-spec                          (5 minutes)

The time allowed for a program started by the server to finish. If this time is exceeded the server sends a TERM signal to the program and then five seconds later a KILL signal.

    ScriptTimeout 10 mins

Equivalent to the NCSA Timeout directive.

ServerRoot dir-name                             (no default)

Specifies the directory that the server uses as the root of the data hierarchy.

    ServerRoot  /usr/public/www

.

Equivalent to the NCSA DocumentRoot directive.

 Meta-information (such as expiry times) about individual documents can be stored in separate files to be added by the server to its HTTP responses. Meta-information files consists of lines of MIME headers, for example:

  Last-Modified: Thursday, 14-Jul-94 15:41:50 GMT
  Expires-by: Monday, 07-Nov-94 12:06:12 GMT

Each file is contained in a subdirectory of the directory containing the document to which it relates. The subdirectory name is specified by the MetaDir directive. The meta-information file has the same base filename as the document to which it relates, but with a suffix specified by the MetaSuffix directive.

MetaSuffix .suffix                                   (.meta)

Specifies the suffix for meta-information files.
There is no equivalent NCSA directive.

MetaDir dir-name                                      (.web)

Specifies the name of a subdirectory in which meta-information files are stored.
There is no equivalent NCSA directive.

Spinning the Web by Andrew Ford
© 1995 International Thomson Publishing
© 2002 Andrew Ford and Ford & Mason Ltd
Note: this HTML document was generated in December 1994 directly from the LaTeX source files using LaTeX2HTML. It was formatted into our standard page layout using the Template Toolkit. The document is mainly of historical interest as obviously many of the sites mentioned have long since disappeared.