Troubleshooting failed OneAgent install

[quangvo-trilogy]

Hi Dynatrace community,

How do I troubleshoot a failed installation of OneAgent on one of my EC2 host?

10:23:00 Updating /proc/sys/kernel/core_pattern with oneagentdumpproc
10:23:49 Starting agents...
10:27:49 oneagent service started
10:27:49 Checking if agent is connected to the server...
10:29:49 Unable to determine Dynatrace OneAgent connection status (oneagentos did not start). For details, see: /var/log/dynatrace/oneagent/installer/installation_554573.log

I don't see anything obvious in the log, here are the last lines:

root@docker-linux-2-dh:~# tail -n 10 /var/log/dynatrace/oneagent/installer/installation_554573.log
2025-12-04 10:27:49 UTC [ERROR] Dec 04 10:27:49 docker-linux-2-dh systemd[1]: oneagent.service: Failed with result 'timeout'.
2025-12-04 10:27:49 UTC [INFO] Successfully executed: systemctl start oneagent-shutdown-detection.service
2025-12-04 10:27:49 UTC [INFO] oneagent service started
2025-12-04 10:27:49 UTC [INFO] Checking if agent is connected to the server...
2025-12-04 10:29:49 UTC [WARN] Command '/opt/dynatrace/oneagent/agent/lib64/oneagentinstallaction --get-start-status' failed, return code: 1, message: Cannot open "/var/lib/dynatrace/oneagent/agent/runtime/startStatus.info" file, error code: 2 (No such file or directory)
2025-12-04 10:29:49 UTC [INFO] Unable to determine Dynatrace OneAgent connection status (oneagentos did not start). For details, see: /var/log/dynatrace/oneagent/installer/installation_554573.log
2025-12-04 10:29:49 UTC [INFO] Cleaning installation temporary files
2025-12-04 10:29:49 UTC [INFO] Removing installation process lock file
2025-12-04 10:29:49 UTC [INFO] Removing installation transaction lock file
2025-12-04 10:29:49 UTC [INFO] Installation finished successfully, PID 554573.

`oneagent.service` started successfully:

root@docker-linux-2-dh:~# systemctl status oneagent.service
● oneagent.service - Dynatrace OneAgent
   Loaded: loaded (/etc/systemd/system/oneagent.service; enabled; vendor preset: enabled)
   Active: active (running) since Thu 2025-12-04 11:16:02 UTC; 5h 22min ago
 Main PID: 106425 (oneagentwatchdo)
    Tasks: 32
   Memory: 1.5G
      CPU: 59min 49.189s
   CGroup: /system.slice/oneagent.service
           ├─106425 /opt/dynatrace/oneagent/agent/lib64/oneagentwatchdog -bg -config=/opt/dynatrace/oneagent/agent/conf/watchdog.conf
           ├─106741 oneagentnetwork -Dcom.compuware.apm.WatchDogTimeout=900 -Dcom.compuware.apm.WatchDogPipe=/var/lib/dynatrace/oneagent/agent/watchdog/oneagentnetwork_pipe_106
           └─106743 oneagentloganalytics -Dcom.compuware.apm.WatchDogTimeout=900 -Dcom.compuware.apm.WatchDogPipe=/var/lib/dynatrace/oneagent/agent/watchdog/oneagentloganalytic

Dec 04 11:13:54 docker-linux-2-dh systemd[1]: Starting Dynatrace OneAgent...
Dec 04 11:16:02 docker-linux-2-dh oneagent[78102]: 11:16:02 Dynatrace OneAgent service started.
Dec 04 11:16:02 docker-linux-2-dh systemd[1]: Started Dynatrace OneAgent.

The thing is, I don't see my host appear in the Classic Hosts app.

Thanks and regards

[AntonPineiro]

Hi,

I would open a support ticket instead of.

Is host able to connect to SaaS server or ActiveGate? It means, TCP flow is open?

Best retards

[DenisL]

Hello, 
Two more suggestions:

1)  Have you ever had the OA installed on that server?.. if so, assure that a full uninstall was performed following documentation (Removing old directories/old reference) > Uninstall OA in Linux >> This is so that the new install does not use old configurations.
>>> Then, run the installation again.

2) Also, another common oversight is enabling OA monitoring in the Host -- so, IF you had this missing host shown in Dynatrace at some point in the past, perhaps you could try to change the timeframe to past 365 days so that it can show in the list of hosts (probably disabled) and you can edit the configuration by changing the toggle to "Monitor this host"

... just another suggestion to check 😉
in case someone else uses this reference
"D"

[quangvo-trilogy]

Thank you all for your help. I'm able to see some metrics after 12 hours, but they are reported intermittently. It looks like my instance is having connection problems to Dynatrace servers. I will keep looking into that direction and open a support ticket later if needed.