Observed tiers

You can define Observed Tiers in the Transaction Flow to expose custom tiers such as DataPower, TIBCO EMS, and load balancer, that would not be visible when no agents are deployed.

Configure observed tier rules

Configure the rules that define Observed Tiers in the Observed tires item of the Edit System Profile dialog box.

From the Transaction Flow or Entry Screen, select Configure Observed Tiers from the context menu.

Observed Tiers can be added or removed after the PurePath data has been captured. Compared to using External API Services, this reduces performance overhead while monitoring the application.

The following columns define an Observed Tier:

  • Criterion: The Observed Tier must match against one of the following selections:

    • Host (Hostname/IP): Matches a hostname or IP address.
    • Server address (Host:Port): Matches a host:port string.
    • Destination (Message): Matches a JMS Topic/Queue name.
  • Match: Select the desired match type. Valid values are equals, starts, ends, or contains.

  • Pattern: Enter the value to match against the selected criterion.

The following columns define how the Observed Tier appears in the Transaction Flow:

  • Service name: Enter a string for the service represented by this Observed Tier. Multiple Observed Tiers with the same service name are aggregated into a single expandable group node in the Transaction Flow.
  • Instance name: Enter a string for a specific instance of the service to allow more fine-grained classification in the Transaction Flow. If several Observed tiers have the same service name but different instance names, expanding that service name's node in the Transaction Flow reveals the individual instance names as child nodes.
  • Vendor type: Select a supported vendor for the Observed Tier from the list. This setting controls the icon that represents this service in the Transaction Flow.
  • Enabled: Set by default. Observed Tiers do not appear in the Transaction Flow unless they are enabled. You can define several Observed Tier rules and then enable or disable them in different combinations to best match your diagnostic needs.

Observed tier appearance

The time booked to an Observed Tier is the sum of all self-times (such as node time without the node's children's times) of all nodes that contribute to that Observed Tier, which is the client-side web request method node's self-time.

Data Power Service (Data Power Instance) showing up as Observed Tier
Data Power Service (Data Power Instance) showing up as Observed Tier