Working with Flows

When you first open the Flows tab, a table lists all flows for the selected server or all servers. For example, the figure below shows the flows for all servers. It also highlights the buttons for stopping and starting flows and clearing statistics.

The table also lists all flows for the selected cluster or all clusters.

If some servers you are monitoring are down and you select a group of servers, such as All, that includes these non-running servers, you may see an informational error message such as the one highlighted in the figure below.

The Flows table includes the following information:

To take action on a flow, select its check box and then click the appropriate button on the right above the table columns. The buttons let you control the flow (start, stop, or pause the flow) and clear the statistics on the flow. If you click the Clear Stats button, a dialog appears asking you to confirm that you want to clear the statistics for the selected flow. You can proceed or cancel the operation.

To view details about a flow, click its name. There are several panes for viewing details about a particular flow accessible via the Summary and Endpoints tabs. These panes are discussed in the next sections.

To view more information about a flow, click the flow name in the table. A Flow Details pane opens and displays a Summary tab with more specific information about the selected flow. A Summary tab display appears for all flows, regardless of the type of the flow. While all Flows regardless of type have an Endpoints tab, the endpoints display may be empty if no endpoints exist. In addition, the Summary tab displays different data and sets of charts depending on the flow type.

Note that you can expand any of the charts on the Flow Details panes. Click the chart itself and it expands. Click the expanded chart to reduce it back to the thumb-nail size. For example, the figure below shows how clicking in the Events Received thumb-nail chart expands it to its full size:

Use the Endpoints tab to see detailed information about a flow’s inbound and outbound endpoints. You can also control endpoints (that is, start or stop a flow’s endpoints) from the Endpoints pane.

You might want to control an endpoint when developing and debugging an application, or when deploying an application. See Use Case Scenarios for Controlling Endpoints.

Click the endpoint(s) you want to control, then select Start Endpoints or Stop Endpoints from the Control Endpoints pull-down list. Endpoints that are currently running are marked with a green circle. Endpoints that are currently not running are marked with a red circle.

Notice that each endpoint, inbound and outbound, has its own table. Identifying and summary information appears first, above the table, followed by the endpoint details in the table itself. The identifying and summary information includes the following:

The table beneath these router totals shows the details for each endpoint using that router. Use the search box to limit the endpoints display to only endpoints whose name matches, or partially matches, the string you enter.

For each endpoint, the details table displays the following:

There are several scenarios when you might want to stop and start an application’s endpoints. This is a particularly useful feature when flows have more than one endpoint.

During development, it may be helpful to stop an endpoint so that you can focus on a particular part of your application. For example, you might want to stop an endpoint so that you can eliminate a potential source of errors. You might do this if you are tracing down an application performance problem. You have determined that something with application messages is slowing down your system, but you need to isolate the source of the message that is causing the bottleneck. In a scenario such as this, you would frequently stop some endpoint, check the performance, then start that endpoint. You would repeat these steps with different endpoints until discovering the problem.

When deploying an application, some circumstances might compel you to verify that your application is not receiving some events while it is still processing certain other events. To make sure that processing is occurring without interference from other messages being received, you might stop corresponding endpoints. After verifying that everything is normal, you would then restart the stopped endpoints. For example, you might have an online shopping website and you want it to focus on web sales over everything else during the holiday season. At these high-volume sales times, you might want to shut down some partner or non-direct requests, and you would so so by stopping certain endpoints.

You might also use this feature when testing a flow that uses multiple endpoints to manage different protocols and requirements. You might want to stop the HTTP external endpoint while you are conducting some tests so that the flow is not exposed to the outside world and thus not used. At the same time, you want to keep other internal endpoints running (these are endpoints within or behind your firewall, such as JMS endpoints) so that you can go forward with your tests.