Request for Node Monitoring

General
, ,
1

Hello,
We want to set up an alert to monitor the active node count in our Pega 8.8.5 application. We need to find out where the “Nodes Running” value displayed in the Admin Studio interface is being fetched from. How can I obtain this value? Is there a specific method for it? I am not sure if the system generates a live log stream for the node count. Alternatively, is there a specific request we can send to get this value as a response?

Although we currently monitor the server processes, sometimes nodes drop from the topology even when the process is still running. We need to detect this issue early. If we can extract this dashboard value from a log or an API request, we can configure our alert mechanism.
Does anyone have any ideas about this, or has anyone worked on it before?

Thanks.

2

In PDC, there is an option for client managed systems to get notifications if a node is down.

But, for Pega Cloud, I did not find any such notification in PDC. We can use D_pzClusterMembersInformation data page to get information about the nodes running in the system.

3

Unfortunately, the link provided for the solution to this issue no longer works. :frowning:

4

This is the latest link. As I said, it only applies for client managed systems but not pega cloud.

5
6

You can use PDC, specifically the topology change

7

Does it also log these codes to Pega_RULES.log on the server?

Our issue is that while the nodes appear as dropped in the interface, the Tomcat process still appears to be running on the server. We want to avoid a delay in detecting that the node has gone down. Maybe we can detect it via the codes from logs.

8

Does the PDC system work with version 8.8.5? Is it an add-on, and where can I find it?

9

yes it does work on version 8.8.5 you can raise a request in pega support for more details

10

Hello again,
It turns out our Pega license doesn’t cover the PDC add-on. Is there a CLI output available for the section of the interface that displays the node count?

11

The “Nodes Running” value in Admin Studio is not generated as a log entry. It is derived from the active node information maintained by the Pega cluster (Hazelcast) and displayed through Admin Studio’s internal APIs. Monitoring only the server process won’t detect this scenario because a node can remain up while being disconnected from the cluster.

Without PDC, you can still monitor active nodes.

Use the Nodes REST API
Call:

GET /api/v1/nodes/all/state

or

GET /api/v1/nodes/{nodeId}/state

The response contains all cluster members, including their running_state, last_heartbeat, host_name, and other node details. Your monitoring tool can count the nodes with running_state = Running and raise an alert if the count drops.

12

There’s no CLI available. You can check with your account executive for getting the PDC add on that would make life easier