Getting Started - Configuring a Simple Veduta™ Example

The below assumes that Veduta is installed and a suitable Java virtual machine is available.

For the purposes of this example we'll configure Veduta to monitor one log file - a Linux system log. However the principles can be carried over to any system log that you wish to monitor.

Configurations

For the initial example you won't need to edit these configurations.

System Configuration - veduta.xml

<?xml version="1.0" encoding="UTF-8"?>
<system>
  <db type="memory">1
    <file>db\db.ser</file>
  </db>
  <network>
    <port>3055</port>
    <server>3055</server>
  </network>
</system>

1. an in-memory db. Veduta can be configured to run in-memory or write its state to a SQL database. For this example we'll run an in-memory database which is backed up to disk (note the file option). See here for more details on choosing database options.
2. a network configuration for the server to listen on (port 3055) and the client to talk to (localhost on port 3055). When moving to another client, this host should be changed to the name of the host the server is running on. See the client/server documentation for more information.

Consumer Configuration - consumers.xml

examples\simple\consumers.xml specifies the configuration thus:

<?xml version="1.0" encoding="UTF-8"?>
<consumers>
  <consumer>
    <file>
      <name>System Messages</name>1
      <filename>examples\simple\logs\messages</filename>2
      <parser type="simple"/>3
    </file>
  </consumer>
</consumers>

1. We specify a name for this configuration
2. We specify the name of the log file
3. We've set a simple parser here. This will read each line and assign a timestamp. As discussed above, this is a simple scenario. We'll show a more useful configuration below.

Report Configuration - report.xml

examples\simple\report.xml specifies the report formats thus:

<?xml version="1.0" encoding="UTF-8"?>

<report>
  <name>Linux system report</name>1
  <description>Reports on occurrences on this Linux system</description>2
  <file>
    <name>examples\simple\results\server</name>3
  </file>

  <consumers>
    <consumer>
      <name>System Messages</name>4
      <label>Host system messages</label>
      <timeslice>5m</timeslice>
      <display>
        <type>chart</type>5
        <markersize>4</markersize>
        <backgroundcolour>#ffffff</backgroundcolour>
        ....
      </display>

      <analysis type="quantity/patterns" values="" patternValues="pattern colour map">6
        <pattern score="1">.*warning.*</pattern>
        <pattern score="2">.*error.*</pattern>
      </analysis>
    </consumer>
  </consumers>

  <times>
    ....
  </times>

  <display>
    ...
    <!-- marker colours -->7
    <values name="pattern colour map">
      <value equals="2" colour="red"/>
      <value equals="1" colour="#ff6600"/> <!-- orange -->
    </values>

  </display>
</report>

1. We specify a report name. This will appear in the generated report.
2. We specify a report description. This will appear in the generated report. This would typically have an explanation of what's being monitored, contact info, hyperlinks etc.
3. We specify a filename for the output reports. This is used as a base name, since timestamps etc. will be appended. Only used if not running Veduta in its webserver mode.
4. We specify a consumer to read information from. This is the consumer that we specified in consumers.xml and the name here must match the name in the consumers.xml configuration.
   Below this we specify the label to be used in the report, how large each time slice displayed will be (here it's 5 minutes)
5. We specify a chart with markers indicating errors etc.
6. We specify how to match strings in the logs and scores to assign for each pattern. These scores are used by the colour mappings below.
7. Finally we specify the colour mappings to use. We can assign different colour sets, thus making colours (or colour combinations) correspond to system activity, messages etc.