Troubleshooting
No Data Troubleshooting
- Check the pod list of the eBPF Agent and the corresponding node.
kubectl -n tingyun get pod -o wide
The eBPF Agent runs in the pod tingyun-linux-agent-xxxxx.
- Confirm whether the eBPF Agent is started on the node.
Run the following command on each node:
kubectl -n tingyun exec -ti tingyun-linux-agent-xxxxx -- ps -ef | grep ebpf_agent | grep -v grep
If the process exists, the eBPF Agent is running. If the process does not exist, the eBPF Agent is not running.
- Confirm the reason why the Agent is not started.
Check whether the kernel version of the node is supported.
kubectl -n tingyun exec -ti tingyun-linux-agent-xxxxx -- cat /proc/version
Compare the system version and kernel version in the support list to confirm whether it is supported.
If the kernel version is within the supported range (kernel version > 4.15.0), check whether the Agent is disabled or in meltdown state.
kubectl -n tingyun exec -ti tingyun-linux-agent-xxxxx -- cat /opt/agent/ebpf/restart.log
- If the Agent is running normally, confirm the Agent working status through the logs.
kubectl -n tingyun exec -ti tingyun-linux-agent-xxxxx -- cat /opt/agent/ebpf/logs/ebpf-agent.log
If data is being uploaded normally, you will see logs similar to the following:
2023-04-21 15:09:49.555 (pid:8815,uid:0) INFO Metrics upload Answer http://tingyun-collector.tingyun.svc:7666/ebpfMetrics?version=1.0.0&license=vZIqa8SUREkDAyMQ
{"status":"success","result":"ok"}
If there is an exception communicating with the Collector, you may see error logs similar to the following:
2023-04-21 15:12:13.602 (pid:20192,uid:0) ERROR http://tingyun-collector.tingyun.svc:7666/ebpfInit?version=1.0.0&license=1234567890 Login Error ConnectFailed
You need to check the Collector deployment status and network status.