Dapper comes with contains two configurable servlets: one provides provides OPeNDAP access to data (referred to as the Dapper servlet), while the other (known as the DChart servlet) provides an AJAX user interface to the OPeNDAP servlet.

Starting or stopping Dapper

Dapper is started with the dapper.sh command followed by the start argument:

    ./dapper.sh start

You must run this command from the Dapper distribution directory.

Debug information from the Tomcat container is available in logs/tomcat.log. Log info from Dapper is sent to logs/dapper.log while log info from DChart is sent to logs/dchart.log.

If you want to run Dapper in a terminal window and have all debug information sent to the terminal, you can use:

    ./dapper.sh run

Dapper is stopped with the stop argument:

    ./dapper.sh stop

Configuring the Dapper servlet

Dapper is configured with the etc/dapper.properties file. A sample configuration file is available in etc/dapper.properties.sample. Following are the configurable properties:

Configuring Dapper aggregations

Dapper allows you to aggregate multiple netCDF files into a single logical netCDF file. This is configured through an aggregation configuration file that is located (by default) in etc/aggs.xml.

Dapper supports two kinds of aggregations:

• Variable aggregations combine the variables from multiple netCDF files into one logical netCDF file
• Outer aggregations concatenate the outer dimension of variables with the same name in multiple files into a logical variable where the outer dimension has a size equal to the sum of the sizes of the outer dimension in the original netCDF files.

Outer aggregations can be nested in variable aggregations. Dapper can aggregate netCDF files that are available as local files as well as by HTTP or OPeNDAP.

Here's a sample aggregation configuration file (available in etc/aggs.xml.sample) that aggregates several netCDF files containing NCAR/NCEP reanalysis data:

 <scichart>
     <aggregation cacheTime="86400" id="www.cdc.noaa.gov/Datasets/ncep.reanalysis.dailyavgs/pressure" type="variable">
         <aggregation type="outer" dim="time">
             <dataset url="http://www.cdc.noaa.gov/Datasets/ncep.reanalysis.dailyavgs/pressure/air.2006.nc"/>
             <dataset url="http://www.cdc.noaa.gov/Datasets/ncep.reanalysis.dailyavgs/pressure/air.2007.nc"/>
         </aggregation>
         <aggregation type="outer" dim="time">
             <dataset url="http://www.cdc.noaa.gov/Datasets/ncep.reanalysis.dailyavgs/pressure/hgt.2006.nc"/>
             <dataset url="http://www.cdc.noaa.gov/Datasets/ncep.reanalysis.dailyavgs/pressure/hgt.2007.nc"/>
         </aggregation>
     </aggregation>
 < /scichart> 

If you copy this file to etc/aggs.xml and restart the server and point your browser to:

http://localhost:8080/dods/aggregations/www.cdc.noaa.gov/Datasets/ncep.reanalysis.dailyavgs/

you will see that these four URls have been consolidated into a single logical OPeNDAP url.

Dapper aggregation file schema

The scichart element is the root element of the Dapper aggregation file.

Dataset to include in an aggregation

Configuring the DChart servlet

DChart is configured with the dchart/etc/config.xml XML file. If the file is missing the servlet will be disabled.

A sample configuration file ( available in dchart/etc/config.xml.sample) is below:

<scichart>
    <title>
    <![CDATA[
	<h1>My Dapper Server</h1>
	<h2>Change me in config.xml</h2>
    ]]>
    </title>
    <property name="opendap.cache.expires" value="3600"/>  <!-- Max cache age in second -->
    <property name="opendap.cache.capacity" value="100000" /> <!-- Max number of items in cache -->
    <datasets>
	<dataset category="Gridded example" name="NCAR Reanalysis" url="file://test/data/agg/air.2006.0.2.nc" type="grid" >
	    <description>
	        Gridded data example
	    </description>
        </dataset>
        <dataset category="In-situ example" url="http://localhost:8080/dods/dapperexample/argo_sample.cdp" name="Argo Pacific profiles">
	    <variables>
	        <include name="PSAL" mapto="Salinity"/>
	        <include name="TEMP" mapto="Temperature"/>
	    </variables>
	    <description>
	        In-situ data example
	    </description>
	</dataset>
    </datasets>
</scichart>

Many more examples are available in the dchart/etc directory, including the configuration file used at PMEL (dchart/etc/config.xml.pmel).

DChart configuration file schema

The scichart element is the root element of the DChart configuration file.

The property element defines name value pairs to pass to the DChart servlet.

Two properties are currently defined:

• opendap.cache.expires specifies how long (in seconds) DChart should cache data from an Opendap server.
• opendap.cache.capacity specifies the maximum number of Opendap requests to store in the DChart Opendap cache.

Contains all of the <dataset> elements and defines the base URL of the Opendap data server.

Use the dataset element to include datasets from the base URL specified by the <datasets> element.

The description element is used to configure a link from a dataset to a description of the dataset. For example:

	    <description>
              <![CDATA[
                More info on this dataset is available <a href="http://www.example.com">here</a>.
              ]]>
	    </description>
		      

Specifies default ranges to use for the dataset. Only time ranges are currently supported.

Specifies a range. Only time ranges are currently supported. Dates are formatted as "DD/MMM/YYYY HH:MM:SS" or [^|$][+/-]delta, where ^ is the first date in the dataset, $ is the last date in the dataset, and delta is a time interval in seconds.All times are assumed to be GMT.

Examples:

<!-- Fixed time interval -->
<ranges>                         
  <range type="t" low="01/Jan/2004 00:00:00" high="31/Dec/2005 00:00:00"/>
</ranges>

<!-- Low value is 1 hour before last time point in dataset -->                        
<ranges>                         
  <range type="t" low="$-3600"/>
</ranges>

Specifies miscellaneous attributes for the dataset.

Name of Dapper variable or attribute that contains the station name. This name will be displayed when a user clicks on a station in the DChart map.

The variables element contains configuration info on the variables in a particular dataset.

Include a variable from a Dapper dataset in DChart dataset and (optionally) rename the variable. For example, if a Dapper dataset contains the variable "PSAL" and you want to display it as "Salinity", you would add:

<include name="PSAL" mapto="Salinity"/>

to the configuration file.

Exclude a variable from a Dapper dataset in DChart.

Specifies the variable to select when a user selects a DChart dataset.

Tomcat configuration

Dapper is packaged with the Jakarta Tomcat Server. By default, the Tomcat server is configured to run on ports 8080 (for HTTP) and 8005 (for shutdown notices). If you already have a server running on either of these ports, you will need to reconfigure Tomcat. The following describes how to change these ports to 9090 and 9005:

• Edit the file jakarta/conf/server.xml in a text editor.
• Change the line containing:

  <Server port="8005"

  to

  <Server port="9005"

• Change the line containing:

  <Connector port="8080"

  to

  <Connector port="9090"

• Save the file and restart the server

More information on configuring Tomcat is available at the Tomcat site.