HTTP ERROR 500 AsyncContext Timeout v1.33.0 and v1.34.0

[Nachtfalkeaw]

Environment

• Telemetry Streaming Version: v1.33.0 and v1.34.0
• BIG-IP Version: BIG-IP 16.1.4.1 Build 0.58.5 Engineering Hotfix
• Resource provisioning: MGMT: Large, LTM: Nominal
• platform C117 - 4 CPUs, 16GB RAM

Summary

I configured an Prometheus Pull Consumer declaration and uploaded it to the API.
I try to pull the metrics and get the "HTTP ERROR 500 AsyncContext Timeout" after around 30s.
ircd_child processes consuming very high CPU and running long after this issue is displayed in the webbrowser.

The error happens always with "ABC_Pull_Consumer" (Default Metrics set by F5, nothing changed) and happens sometimes with "ABC_Pull_Consumer_2". This is on the standby system no load from User Traffic.

{
	"class": "Telemetry",
	"controls": {
         "class":"Controls",
         "logLevel": "info",
		 "debug": false
     },
    "ABC_System": {
        "class": "Telemetry_System",
        "enable": "true",
        "systemPoller": [ "ABC_System_Poller1", "ABC_System_Poller2" ]
    },
	"ABC_System_Poller1": {
        "class": "Telemetry_System_Poller",
		"trace": false,
		"interval": 0,
		"enable": true,
        "host": "localhost",
        "port": 8100,
        "protocol": "http",
        "allowSelfSignedCert": true
	},
	"ABC_System_Poller2": {
		"class": "Telemetry_System_Poller",
		"trace": false,
        "interval": 0,
        "host": "localhost",
        "port": 8100,
        "protocol": "http",
        "allowSelfSignedCert": true,
		"endpointList": [ "Endpoints_ABC" ],
		"enable": true
	},
	"Endpoints_ABC": {
		"class": "Telemetry_Endpoints",
		"items": {
			"profileHttpStats": {
                 "path": "/mgmt/tm/ltm/profile/http/stats"
             },
			"system_performanceConnections": {
				"path": "/mgmt/tm/sys/performance/connections/stats?options=detail"
			},
			"system_tmmTraffic": {
				"path": "/mgmt/tm/sys/tmm-traffic"
			},
			"system_performanceThroughput": {
				"path": "/mgmt/tm/sys/performance/throughput/stats?options=detail"
			},
			"nodeStats": {
				"path": "/mgmt/tm/ltm/node/stats"
			},
			"node": {
				"path": "/mgmt/tm/ltm/node"
			}
		}
	},
	"ABC_Pull_Consumer": {
		"class": "Telemetry_Pull_Consumer",
		"trace": false,
		"type": "Prometheus",
		"enable": true,
		"systemPoller": [ "ABC_System_Poller1" ]
	},
		"ABC_Pull_Consumer_2": {
		"class": "Telemetry_Pull_Consumer",
		"trace": false,
		"type": "Prometheus",
		"enable": true,
		"systemPoller": [ "ABC_System_Poller2" ]
	}
}

Steps To Reproduce

Upload declaration, Access the Pull_Consumer Endpoints using webbrowser (or Prometheus)

Error Log

/var/log/restjavad.0.log shows this error:

[WARNING][5296][18 Jan 2024 08:51:26 UTC][8100/shared/iapp/build-package BuildRpmTaskCollectionWorker] Failed to execute the build command 'rpmbuild -bb --define '_tmppath /var/system/tmp' --define 'main /var/config/rest/iapps/f5-telemetry' --define '_topdir /var/config/rest/node/tmp' '/var/config/rest/node/tmp/8fb6a962-332b-4029-adcf-02f29cfd1e60.spec'',  Threw:com.f5.rest.workers.shell.CommandExecuteException: Command execution process killed
        at com.f5.rest.workers.shell.ShellExecutor.finishExecution(ShellExecutor.java:282)
        at com.f5.rest.workers.shell.ShellExecutor.access$000(ShellExecutor.java:34)
        at com.f5.rest.workers.shell.ShellExecutor$1.onProcessFailed(ShellExecutor.java:321)
        at org.apache.commons.exec.DefaultExecutor$1.run(DefaultExecutor.java:203)
        at java.lang.Thread.run(Thread.java:748)

Expected Behavior

OpenTelemetry Plugin should provide the metrics in Prometheus format within a few seconds without consuming all CPU.

Actual Behavior

No metrics, high CPU for long time (5-10 minutes)

[RahmanBadru]

any fix here