1. Service Availability Widget

1.1. Purpose

The purpose of the Service Availability widget is to monitor the success rate of defined connections during predefined periods of time. Note that this widget only works for one single Engine. If one'd like to monitor connections on several Engines, one has to create one widget per Engine.

1.2. Creating widget

1. select on which Engines the widget will be computed

   
   

2. configure the connections you want to monitor (called "services") along with the time period and a target
3. optionally group the results using a category

   
   

4. enter a name to identify the service
5. set hours on which the analysis will be done. Let 00:00 means the full day
6. set days on which the analysis will be done
7. set the target. It correspond to the ratio of successful connections on total connections
8. optionally add conditions on objects

   
   

Note that by default a week begins on Monday and ends on Friday. It implies that if the time period is configured on weekdays then Saturday and Sunday will be ignored. The days of a week could be configured using the configuration file of the Portal. This configuration file is named portal-general-config.xml and is in conf directory. The syntax is the following:

<!--each el is a working day of the week and the number correspond to a day
    0 = Sunday, 1 = Monday, ..., 6 = Saturday -->
<weekDays>
  <el>1</el>
  <el>2</el>
  <el>3</el>
  <el>4</el>
  <el>5</el>
</weekDays>

1.3. Using widget

The results are summarized in a table and can be viewed with a "weekly" or "monthly" granularity.

If results are grouped then values could be displayed per keyword using the list in the footer

Day by day values can be obtained by clicking on the "details" button.

It is important to understand how the number of users is computed. On each source there are some users managed by the system. These users are tagged as 'system' by the Engine but there is only one user by type (Network Service, Local Service, Local System) instead of one by source. It means that when a service is used by a system user the computed number correspond to the number of sources used for the connections. Normal users are counted as 1.