Uploaded image for project: 'SAFe Program'
  1. SAFe Program
  2. SP-3895

Taranta captures user attention when things are not working

Change Owns to Parent OfsSet start and due date...
    XporterXMLWordPrintable

Details

    • Obs Mgt & Controls
    • Hide

      Providing better situational awareness to the user of running dashboards when there are connection problems with underlying devices. The user will quickly notice that something is bad behaving making the dashboard unreliable.

      Show
      Providing better situational awareness to the user of running dashboards when there are connection problems with underlying devices. The user will quickly notice that something is bad behaving making the dashboard unreliable.
    • Hide

      See the 3 incremental solutions below:

      1. configurable fail-safe behaviour
      2. global reliability indicator
      3. detailed info of failures
      Show
      See the 3 incremental solutions below: configurable fail-safe behaviour global reliability indicator detailed info of failures
    • 6
    • 2
    • 0
    • Team_CREAM
    • Sprint 5
    • PI24 - UNCOVERED

    • Taranta Team_CREAM

    Description

      (this comes from a discussion raised by s.vrcic on slack: https://skao.slack.com/archives/C012S5JFTPE/p1700149373310919)

      We need to make Taranta capable of letting the user know when in a running dashboard some anomalies happen. In particular, 

      • when a device attribute does not respond anymore;
      • when a device attribute changes so fast that somehow updates are lost by Taranta;
      • when a device attribute produces data that is not valid.

      From a UI perspective what we could do is:

      • add a status bar to Taranta, when a dashboard is running, with a pulsating LED
      • the LED is green and pulses when data is coming as expected from all devices linked to the dashboard
      • the LED is orange and pulses when some device is sending data, but some data are invalid;
      • the LED does not pulse and is red when some device looses connection.
      • in addition to color, there's also some icon that changes to reflect these changes in the connection health (for accessibility purposes);
      • if the user clicks on the LED a popup dialog shows details (green --> everything is fine; orange --> which devices are sending bad data; red --> which devices are lost).

      As Sonja said in her comment below, it is also important to inform the user of which devices are disconnected or providing invalid data.

      Original message by s.vrcic on Nov 16.

      Hello Cream Team, I have a question related to how faults are reported on the Taranta dashboards: # if a Taranta dashboard fails to  subscribe for one of the attributes  how is this reported by the dashboard?  (For example  the  FQDN has been misspelled, or the version of the Tango Device does not implement the attribute. )Can the dashboard designer select how is this reported (colour of the background or the frame of the filed, or something else).

      1. If the Tango Device implements the attribute, but  is not able to obtain the value from the 'actual' device (device not responsive or query fails for any reason),  how is this reported by the Tango device and on the Taranta Dashboard?  

      Do Taranta widgets allow developers to select how the Attribute Quality is interpreted  and used to change colour of the background or text, or to display an error message.  Can you point me to the  relevant documentation ?

      Change in scope occurred during PI22 planning

      Based on the comment below and extensive discussion with CREAM we decided to descope as follows.

      To spread this work on 2 layers:

      1. explicitly configuring the 'fail-safe' behavior of Taranta (that is automatic and that might mask some underlying malfunctions in Tango devices)
      2. providing the user of a running dashboard with a global 'reliability' indicator that says if everything if fine, or not

      Explicitly configuring the fail-safe behaviour:

      1. right now TangoGQL uses 3 mechanisms to interact with tango devices: event subscription, periodic 5s polling, on-demand polling. It automatically switches between them (in the order given above) when one mechanism appears not to be working. This behaviour, which aims at improving reliability, in effect in the past caused fault masking, making it more difficult for people to understand why a dashboard was not behaving as it should have.
      2. This is expecially critical for some of the widgets, like the Timeline, that rely on events being signalled by devices.
      3. We decided that we want to enable the capability (for the dashboard designer) to specify, when configuring the widget, if to enable or not this fail-safe behaviour. In this way, the timeline widget when for some reason the tango device fails to emit events will 'know' that something is wrong and make it clear to the user.
      4. This configuration flag will be added to all widget types that are event-driven.

      Global reliability indicator

      Basically what is described above as a pulsing LED, with 3 states:

      1. slow pulsing (eg 0.1Hz), green: to signal that everything is fine
      2. faster pulsing (eg 0.5Hz), orange:  to signal that some device has some attribute that is 'INVALID' or 'DEGRADED' (which, according to tango, means that some data of that attribute was lost)
      3.  even faster flashing (eg 1Hz), red: to signal that some device is not connected to Taranta.

      The descoping affects the capability to display what is not working, which will not be tackled in PI22.

      Attachments

        Structure

          Activity

            People

              g.brajnik Brajnik, Giorgio
              g.brajnik Brajnik, Giorgio
              Votes:
              0 Vote for this issue
              Watchers:
              1 Start watching this issue

              Feature Progress

                Story Point Burn-up: (100.00%)

                Feature Estimate: 6.0

                IssuesStory Points
                To Do00.0
                In Progress   00.0
                Complete918.5
                Total918.5

                Dates

                  Created:
                  Updated:

                  Structure Helper Panel