Use the monitoring interface

Access the monitoring user interface

To access the monitoring user interface, on the main menu, click the Security & Monitoring  icon. Then access any of the menu options under Monitoring

Users need the Access platform management tools permission to access the monitoring user interface.

If the Monitoring tab is not displayed on the left menu, it means that the currently logged user lacks this permission.

The following describes the different tab menus under Monitoring and what they enable you to do. 

Current activity tab

The Current Activity  tab displays the current activity of the deployment.

This screen is automatically refreshed every 5 seconds. To force a refresh, click the Refresh button. 

Node usage area

This section displays information about the overall health of Axway Decision Insight Primary or Standalone Node. The area comprises the following elements:

  • A CPU counter displays the average CPU usage of the system (including all processes that runs on the system) during the past 5 seconds. 
  • A CPU chart displays the evolution of the CPU usage of the past minute.
  • A memory counter displays the current usage of the JVM heap with a focus on the old generation memory pool usage.

If the memory usage is high, click the Reclaim memory button to trigger a garbage collection. If no memory is freed in the next few minutes, it means that DI is fully using the allocated memory.

Primary/replica cluster usage area

When you're connecting to the primary node of a primary/replica cluster, this area provides information about the cluster topology:

  • The Replica column shows the  IP or the hostname of your replica nodes (RN) – provided the hostname can be resolved from the primary node. 
  • The Type column shows whether an RN is dedicated to dashboard queries only, or if distributed computings are enabled on it.
  • The Status column shows the status of the RNs in the cluster in relation with the primary node:
    • OK  – the RN is ready to handle dashboard requests.
    • LATE  – the RN is catching up the PN (the delay is displayed after the status).
    • OFFLINE – the RN is useless to the cluster (even if it is running and trying to connect to the primary node).
    • INITIALIZING – the RN is initializing (by initializing its database, restoring its checkpoint). While checkpoint restoration is processing, the following indicators are available:
      • The progression of the checkpoint download
      • An estimation of the remaining time (depending on the network bandwidth)
      • The elapsed time
      • The amount of checkpoint files downloaded, over the amount of all the files to download
      • The size of all checkpoint files downloaded, over the size of all the files  to download
  • The Network column shows the amount of bytes, that the replica node exchanged with the primary node (over the last 10 seconds, with a shift of 10 seconds).

For more information about primary/replica clusters, see Clustered deployment.

Attribute computing area

This area displays information about the attributes that are being computed.

Most attributes are computed in a few milliseconds and attributes that are listed here might have been computed since the moment you first opened the tab.

Attribute precomputing area

This section displays information about precomputings currently in execution.

Data integration area

This section displays information about the mappings that are being absorbed. Most mappings are absorbed in a few milliseconds and mappings that are listed here might have finished since the moment you first opened the tab.

Dashboard queries area

This section displays information about the dashboard queries that are being executed.

Most queries are executed in a few milliseconds and queries that are listed might have finished since the moment you first opened the tab. In case a query takes far too long to be executed, you can abort it using the Cancel button.

Activity report tab

The Activity Report tab displays the past activity of DI. The information displayed is the same as the information in the Current tab.

Computing tab

This tab displays a computing monitoring console which lists the attributes being computed along with additional information.

Hide idle attributes area

At the top of the console, select this check box to display only the attributes which are live at the moment, that is attributes with some tasks waiting to be computed.

Filter attribute list

Use the search box to the right of the Hide idle attributes check box to filter attributes based on attribute name and entity. 

Console columns

The following table describes what type of additional information you can view from the columns of the console.

Information Details

Status

  • Live – Attribute is being computed in real time. When a calculation is required, it is done as soon as possible.
  • Catch up – Attribute is being computed in the past.
  • Paused – Computing for this attribute has been paused.
Computed intervals

The interval during which the attribute has been computed.

If computing intervals are disjoined:

  • The beginning of the first interval and the end of the last interval are displayed.
  • A tooltip provides more information and displays the first three intervals and the last one.
Live Number of tasks waiting to be computed to fill attribute values for the first time.
Correction

Number of tasks waiting to be computed for a requested correction. A task usually represents an interval of one hour.

A correction is triggered when an attribute has already been computed on an interval and some information has changed.

Precompute
  • Precompute – Forces the precomputation of an attribute that is eligible for precomputing and that is not already precomputed.
  • Stop – Force-stops the precomputing of an attribute that is currently precomputed.
  • N/A – Means that the attribute is not eligible for precomputing, so no pre-computations actions are available.

The actions available on attributes can have a direct impact on performance and coherency. Use them carefully.

Actions
  • Pause – Force-stops computing an attribute until further notice.
  • Resume  Allows continuing computing a paused attribute.
  • Cancel –  Destroys all the tasks that are waiting to be computed.
  • Recompute – Triggers attribute computing from a given date.

The actions available on attributes can have a direct impact on deployment performance and coherency. Use them carefully.

Pending corrections area

To prevent high correction overhead, the collected "late data" (data injected in the past) is not computed immediately.
Their computing interval is stored in a pending queue and flushed automatically (by default, after 15 minutes or 100 000 events).

Actions

Available actions are:

  • Flush all will trigger corrections over late data intervals immediately.
  • Drop all will drop all late data (corrections are forgotten).

Filter attribute list

Use the search box to filter attributes based on attribute name and entity. 

Precomputing tab

The Precomputing tab displays the activity and status of precomputings on DI.


Related Links