{section}
{column}

Two kinds of graph can be generated by Cacti:
* using a simple data: data are returned by the probe and are directly and completely graphed,
* or using a data query: data are returned by the probe and must be filtered before to be graphed

{note}The following steps by steps process must be adapted to your metric{note}

{column}
{column:width=350px}
{panel:title=Table of contents}{toc:outline=true}{panel}
{panel:title=Contributors}{contributors:order=name|mode=list|showAnonymous=true|showCount=true|showLastTime=true}{panel}
{column}
{section}

{anchor:Creating a graph using a simple data}

Cacti monitoring extension of Petals CLI is used to get metric values. See the documentation of [Petals CLI|petalscomponents:Petals CLI] for more information.
In the *Input String*, you will put the command line arguments to connect to the the Petals node for which you want to graph the delivered messages of the local transporter. That's why the command line contains: *<petals_host>*, *<petals_jmx_port>*, *<petals_jmx_user>* and *<petals_jmx_password>*.
# Add an *input field* for your Petals node hostname
!petals-container-localtransporter-dim-petals-host.jpg|align=center!
As the Petals node hostname can be automatically set from the hostname of the monitored host, the field *Special Type Code* is set.
# Add others *input fields*
!petals-container-localtransporter-dim-petals-jmx-port.jpg|align=center!
As for the JMX port, create an input field for the Petals JMX username and the Petals JMX user password.
# Add the *output fields* returned by the script
For each output fields of the script/command (here, in the sample, for each message exchange status "Active", "Done" and "Error"), create an output field:
!petals-container-localtransporter-dim-output-active.jpg|align=center!
# Check that you have the following *Data Input Method*:
!petals-container-localtransporter-dim-full.jpg|align=center!
# Create a new *Data Template*:
!petals-container-localtransporter-dt.jpg|align=center!
Select your data input method previously created as *Data Input Method*.
Add your first *data item* associated to an output field of the script/command previously created (here in the sample, the number of delivered messages for message exchange status 'Active'). The *data source type* of the item must be correctly set according to the output field type.
# Create the others *data items* associated to the remaining output fields
!petals-container-localtransporter-dt-items.jpg|align=center!
# A new block *Custom data* appears, check *Use Per-Data Source Value* for all except for the Petals node hostname
!petals-container-localtransporter-dt-custom.jpg|align=center!
# Create a new *Graph Template*:
!petals-container-localtransporter-gt.jpg|align=center!
# Add a *Graph Item* to graphed one of the previous *data item* (here in the sample, the number of delivered messages for message exchange status 'Active')
!petals-container-localtransporter-gt-active.jpg|align=center!
# Add a *Graph Item* to displayed the legend of your *data item* (here in the sample, the number of delivered messages for message exchange status 'Active')
!petals-container-localtransporter-gt-active-legend.jpg|align=center!
# Add others *Graph Items* associated to the remaining *data items*
!petals-container-localtransporter-gt-full.jpg|align=center!
# Create a *New Graph*
!petals-container-localtransporter-ng.jpg|align=center!
# Set parameters needed to the *Data Input Method*
!petals-container-localtransporter-ng-params.jpg|align=center!

The new graph can be viewed after several minutes needed to collect data. By default a collect is done every 5 minutes. To view the graph, go into *Graph management"

{anchor:Creating a graph using a data query}
h1. Using a data query

<name>Get Delivered Messages by the Local transporter to an endpoint</name>
<script_path>/home/cdeneux/petals-cli-2.1.1-SNAPSHOT/bin/petals-cli.sh</script_path>
<arg_prepend>-h localhost -n 7700 -u petals -p petals -c -- monitoring -o local-transporter -f delivered-messages</arg_prepend>
<arg_index>-- --list-endpoints</arg_index>
<arg_query>-- --query-endpoints</arg_query>
<arg_get>-- --query-endpoints</arg_get>
<output_delimeter>!</output_delimeter>
<index_order>endpointName</index_order>
<index_order_type>alphanumeric</index_order_type>
<!--index_title_format>|chosen_order_field|</index_title_format-->

<fields>
<endpointName>
<name>Endpoint name</name>
<direction>input</direction>
<query_name>NAME</query_name>
</endpointName>
<activeDeliveredMsgs>
<name>Active Message exchange delivered counter</name>
<direction>output</direction>
<query_name>ACTIVE</query_name>
</activeDeliveredMsgs>
<doneDeliveredMsgs>
<name>Done Message exchange delivered counter</name>
<direction>output</direction>
<query_name>DONE</query_name>
</doneDeliveredMsgs>
<errorDeliveredMsgs>
<name>Error Message exchange delivered counter</name>
<direction>output</direction>
<query_name>ERROR</query_name>
</errorDeliveredMsgs>
</fields>
</interface>
{code}
Don't forget to adapt the path of Petals CLI to your installation. Petals CLI is used to collect data. See the documentation of [Petals CLI|petalscomponents:Petals CLI] for more information.

To create this graph, please follow the following steps:
# First, create a new *Data Query*:
!petals-container-localtransporter-itf-dq.jpg|align=center!
# Go to the *device* for which you want collect metric values, and associate it to the data query previously created
!petals-container-localtransporter-itf-device-associate.jpg|align=center!
# Create a new *Data Template*:
!petals-container-localtransporter-itf-dt.jpg|align=center!
Select "*Get Script Data (Indexed)*" as *Data Input Method*.
Add your first *data item* associated to an output of the data query (here in the sample, the number of delivered messages for message exchange status 'Active'). The *data source type* of the item must be correctly set according to the output field type.
# Create others *data items* associated to the remaining output fields (here in the sample, number of delivered messages for message exchange status 'Done' and 'Error')
!petals-container-localtransporter-itf-dt-items.jpg|align=center!
# A new block *Custom data* appears, check *Use Per-Data Source Value* for all
!petals-container-localtransporter-itf-dt-custom.jpg|align=center!
# Create a new *Graph Template*:
!petals-container-localtransporter-itf-gt.jpg|align=center!
# Add a *Graph Item* to graphed one of the previous *data item* (here in the sample, the number of delivered messages for message exchange status 'Active')
!petals-container-localtransporter-itf-gt-active.jpg|align=center!
# Add a *Graph Item* to displayed the legend of your *data item* (here in the sample, the number of delivered messages for message exchange status 'Active')
!petals-container-localtransporter-itf-gt-active-legend.jpg|align=center!
# Add others *Graph Items* associated to the remaining *data items*
!petals-container-localtransporter-itf-gt-full.jpg|align=center!
# Go to the *data query* previoulsy created, and associate it to the just create graph template
!petals-container-localtransporter-itf-dq-associate.jpg|align=center!
# Associate outputs of the *data query* to the *graph items*
!petals-container-localtransporter-itf-dq-associate-output.jpg|align=center!
*Don't forget to check boxes.*
# Go to the *device* for which you want to create the graph, and *Create Graphs for this Host*
!petals-container-localtransporter-itf-ng.jpg|align=center!
Select yours endpoints for which you want create a graph

The new graph can be viewed after several minutes needed to collect data. By default a collect is done every 5 minutes. To view the graph, go into *Graph management"