cdp-doctor fluentd status

Scope

The cdp-doctor fluentd status command returns a detailed runtime snapshot of the logging agent processes and network buffers on the node.

It reports service health, network access, process PIDs, and resource usage, and buffer sizes — useful for operational troubleshooting and capacity checks.

Service status and Salt configuration flags.
Network access to S3, STS, and Databus endpoints.
Supervisor and Agent process details (PID, CPU, memory, VMS, RSS, open files).
Size of Fluentd buffers (per stream/buffer), indicating backlog in local disk buffers.

Use Case

When log shipping is delayed/missing in the Control Plane.
Confirming agent process resource usage (memory, open files).
After reinitializing or reconfiguring Fluentd (fluent.init) to validate the runtime state.
Checking configuration and connectivity health for the logging agent.
Creating runtime/process snapshot (PIDs, resource use, buffers) for operational troubleshooting.

Sample Output

Running the cdp-doctor fluentd status command displays the following output:

Logging agent service details:
+-------------------+------+
| cdp-logging-agent | [OK] |
|  salt configured  | [OK] |
+-------------------+------+
Network checks:
+---------------------------+------+
|       S3 Accessible       | [OK] |
|      STS Accessible       | [OK] |
|  Databus API Accessible   | [OK] |
| Databus S3 API Accessible | [OK] |
+---------------------------+------+
Logging process 'Supervisor':
PID: 15583  CPU: 0.0  RSS: 41.9 MB  Open files: 1
Logging process 'Agent #1':
PID: 15588  CPU: 1.4  RSS: 172.5 MB  Open files: 714
Size of buffers:
s3_CM_COMMAND : 507.0 B
s3            : 235.8 KB

Service [OK] and network [OK] results indicate that the agent is healthy and can ship logs.
A large open file count (e.g., 714) for Agent process RSS/Open files is normal for heavy log forwarding, but should be monitored.
Small Buffer sizes indicate no backlog. Large or increasing buffer sizes indicate shipping delays or endpoint throttling.