RobotLimit

Specify the maximum number of pages a user can visit without a short pause. Also sets a limit on the number of unique sessions that can be created for a single IP address over a given period of time.

Synopsis

RobotLimit number

The parameter is expected to be a positive integer. The default is zero (no limit).

Scope

This directive is only available for use in the local (catalog.cfg) configuration file. It will not affect any other website in any way. This directive will not work in the global (interchange.cfg) configuration file.

Description

This directive defines the number of consecutive pages a user session may access without a 30 second pause. If the limit is exceeded then the following will happen:

The command defined in the LockoutCommand global directive will be executed.
Further website URIs will be generated using "http://127.0.0.1/" as the VendURL, sending the bad robot back to itself when it follows links.

The above-mentioned 30 second pause can be configured using the Limit directive's "lockout_reset_seconds" setting.

If a remote client gets assigned a new session with every visit, rather than maintaining a single session, then this directive sets the limit for the number of sessions that may be assigned to the individual IP address over a one hour period. The one hour period can be overridden using the Limit directive's "ip_session_expire" setting (value specified in minutes). If a lockout is triggered it will remain in effect for 24 hours. The 24 hour period can be overridden with the Limit directive's "robot_expire" setting, as shown below.

Availability

In Interchange versions prior to 5.5.0, this directive controlled the number of sessions allowed for a single IP address without a 1 day pause. The 1 day pause could be overridden with the Limit directive's "robot_expire" setting.

The behaviour described on this page is correct for Interchange versions 5.5.0 and later.

Compatibility Note

When upgrading from an Interchange version prior to 5.5.0, you should delete all of the session per IP counter files in the "tmp/addr_ctr" directory, under your website's home directory (assuming ScratchDir is set to "tmp"). The website's home directory is specified using the Catalog global configuration directive. See the "UPGRADE" file for more details.

If a "lockout" SpecialSub is defined then that subroutine will be called to determine whether the lockout should be overridden for this user. If the "lockout" SpecialSub returns a true value then the current user will not be locked out of your website.

Examples

Setting the "bad robot" limit

RobotLimit 100

Reducing the required pause time

Limit lockout_reset_seconds 20

Overriding the default 24-hour lockout time

Limit robot_expire 0.04

The above will reduce the lockout time from 24 hours to just under one hour.

Adding a hook to determine whether a lockout should occur at all

GlobalSub lockout_check <<EOS
sub {
return $::Scratch->{override_lockout};
}
EOS

SpecialSub lockout lockout_check

The above code will return the value of the "override_lockout" scratchpad variable. If the value is true (defined value that is not zero and not blank) then the user will not be locked out for exceeding the RobotLimit.