HERITRIX_HOME is used throughout this document to refer to the root of the installation.
The startup scripts use the environment variable
HERITRIX_HOME. You can set it manually, but normally the startup scripts are able to figure it out themselves.
Here is the simplest way to start heritrix:
$HERITRIX_HOME/bin/heritrix --webui-admin PASSWORD
This starts the heritrix crawl engine and web UI in the background, logging to
$HERITRIX_HOME/heritrix_out.log, listening on http://localhost:8080/ with the specified password. Remote JMX support may also be initialized (see the section Remote JMX Configuration below for more information).
There are more examples at the bottom.
Foreground vs. Background
There are two Heritrix launcher scripts:
Typically you will use
$HERITRIX_HOME/bin/heritrix, which forks heritrix into the background and logs to an external file,
$HERITRIX_HOME/heritrix_out.log by default.
Useful for debugging,
$HERITRIX_HOME/bin/foreground_heritrix does not fork and logs to stdout.
Command Line Options
$ ./heritrix --help
-a,--webui-admin <arg> Specifies the authorization password
which must be supplied to access the webui. Required if launching the
-b,--webui-bind-hosts <arg> A comma-separated list of hostnames for
the webui to bind to.
-h,--help <arg> Usage information.
-j,--jobs-dir <arg> The jobs directory. Defaults to ./jobs
-k,--keystore <arg> Path to keystore with private key and
certificate for webui running behind https.
-l,--logging-properties <arg> The full path to the logging properties
file (eg, conf/logging.properties). If present, this file will be used to
configure Java logging. Defaults to ./conf/logging.properties
-n,--no-web-ui Do not run the admin web user interface;
only run the crawl engine. If set, the crawl engine will need to be
controlled via JMX or a remote web UI.
-p,--webui-port <arg> The port the webui should listen on.
Defaults to 8443 with SSL enabled or 8080 otherwise.
-r,--run-job <arg> Specify a ready job or a profile name to
launch at launch. If you specify a profile name, the profile will first
be copied to a new ready job, and that ready job will be launched.
-u,--no-engine Do not run the crawl engine; only run the
admin web UI.
-w,--webui-war-path <arg> The path to the Heritrix webui WAR.
-x,--no-ssl Use plain http and not https for the
Web UI and/or Crawl Engine
The web-based user interface and the crawl engine are distinct pieces of heritrix. Typically they run together, but that is not required.
Use this option to run only the web UI. Once heritrix is launched, you can hook up crawl engines through the web UI.
Use this option to run only the crawl engine. The engine can then be controlled through remote JMX or through a separate web UI. See below for more info on JMX.
Web UI Configuratioin
Sets the admin password. This option is required, there is no default.
This option takes a comma-separated list of hosts/ip addresses to listen on (though typically one is enough). Specify
0.0.0.0 to listen on all interfaces.
The default is to listen only on localhost.
For security, when this option is specified, the web UI will be served over https (SSL encrypted), unless --no-ssl is specified.
If SSL (https) is enabled, the default port is 8443, otherwise it is 8080. Use this option to specify a different port.
By default heritrix will look for the WAR in
$HERITRIX_HOME/lib. Use this option to specify a different location. The path can point to either a WAR file or an exploded webapp directory.
Use this option to force the web UI to be served over plain, insecure http. Use of this option is discouraged.
By default, when SSL is enabled, heritrix will generate a private key and certificate at startup.
For a variety of reasons, particularly if you are running heritrix as a public-facing service, you may want to use your own certificate. To do this, you can create a keystore using the java keytool. There is some rudimentary information on using keytool here: http://docs.codehaus.org/display/JETTY/How+to+configure+SSL. Note: the password on the keystore and on the key must be "heritrix".
Unless there is a compelling rationale, use of a keystore is discouraged. It is more secure to let heritrix generate a key/cert on the fly, since there is no secret file that can be compromised.
The default jobs directory is
$HERITRIX_HOME/jobs. Use this option to specify a different path.
By default, heritrix will look for logging configuration in
$HERITRIX_HOME/conf/logging.properties. Use this option to specify a different path. There is some information on customizing this file in the one that comes with heritrix.
Start the specified ready job immediately. The argument is the name of the job, which is the name of the subdirectory, not including the prefix "
ready-". See an example
HERITRIX_HOME sets the root of the installation. Search this document for "HERITRIX_HOME" and look at each instance to discover what this variable affects concretely.
Heritrix (the startup script) writes the process id in
$HERITRIX_HOME/heritrix.pid. It does not make further use of this file, does not complain if it is overwritten, and does not delete even after shutting down.
The default log location is
HERITRIX_OUT to specify a different location.
java command is on the system path, the startup script will set
JAVA_HOME based on that. Use this environment variable to specify a different location, or if
java is not on the path.
Remote JMX Configuration
The crawl engine can be remotely controlled via JMX.
--no-webui is specified, remote JMX must be enabled and configured properly, otherwise the crawl engine would be inaccessible.
A password file is required for remote JMX, and it will start automatically if the password file exists. See
$HERITRIX_HOME/conf/jmxremote.password.template for more information.
The default JMX port is 8849. Use this environment variable to specify a different port.
Set this variable to a non-empty string, e.g.
JMX_OFF=off, to disable remote JMX.
$ heritrix --webui-admin heritrix-password
WARNING: $HERITRIX_HOME/conf/jmxremote.password not found.
WARNING: Disabling remote JMX.
Fri Nov 21 00:05:46 PST 2008 Starting heritrix...
No JNDI context.
Engine registered at org.archive.crawler:instance=14434757,jmxport=-1,name=Engine,type=org.archive.crawler.framework.Engine,host=noah-levitts-computer.local
Web UI listening on localhost:8080.
- logs to $HERITRIX_HOME/heritrix_out.log
- web UI on http://localhost:8080/
- password "heritrix-password"
- JMX on port 8849 if $HERITRIX_HOME/conf/jmxremote.password exists
$ HERITRIX_HOME=/var/heritrix HERITRIX_OUT=/var/log/heritrix.log JMX_PORT=9949 JAVA_HOME=/usr/lib/jvm/jdk1.6.0_06 heritrix -a heritrix-password --webui-bind-hosts 0.0.0.0 -j /var/heritrix-jobs
Fri Nov 21 00:33:47 PST 2008 Starting heritrix..
No JNDI context.
Engine registered at org.archive.crawler:instance=30931963,jmxport=9949,name=Engine,type=org.archive.crawler.framework.Engine,host=cheese
Web UI listening on 0.0.0.0:8443.
$ ./heritrix -a admin --run-job broad_but_shallow-20081121085437
Fri Nov 21 00:55:18 PST 2008 Starting heritrix.....
No JNDI context.
Engine registered at org.archive.crawler:instance=7832149,jmxport=8849,name=Engine,type=org.archive.crawler.framework.Engine,host=noah-levitts-computer.local
Web UI listening on localhost:8080.
- This runs the job in
$HERITRIX_HOME/jobs/ready-broad_but_shallow-20081121085437, which would normally be initialized through the web UI, but could be mimicked outside heritrix.
Heritrix can be shut down cleanly by killing the process with the default signal SIGTERM. It can also be shut down throught the web UI.