In this article, we will explore how to troubleshoot EVE-NG nodes that fail to start properly. The symptom is straightforward yet frustrating: when attempting to start a node, it appears to launch but then shuts down after a few seconds without any explanation.

Unfortunately, some architectural decisions made early on have made it challenging to surface these types of backend errors to the user. This was a deliberate choice, based on the assumption that with correct usage, such failures would be extremely rare. This assumption held true for the most part, but in those few error cases, users remain helpless.

Here, we will walk through the correct troubleshooting steps. We will connect to the EVE-NG system via SSH as the root user and manually replicate the backend process to identify the failure point.

Logs

Logs are located in /opt/unetlab/data/Logs (note that UNetLab was the predecessor of EVE-NG, and much of the original structure remains intact). The primary file of interest is unl_wrapper.txt, which records all operations performed by the wrapper, the program that mediates communication between the node and the backend. Specifically, the wrapper manages node start, stop, and console operations.

Run the following command:

tail -f /opt/unetlab/data/Logs/unl_wrapper.txt

Start the problematic node and observe the output. Key lines to note include:

Aug 01 10:07:03 INFO: CWD is /opt/unetlab/tmp/5/1fdf27c0-5e97-4706-a9da-1a82b01c0fde/2
Aug 01 10:07:03 INFO: starting /opt/unetlab/wrappers/qemu_wrapper -C 54179 -T 5 -D 2 -t "vIOS" -F /opt/qemu-2.4.0/bin/qemu-system-x86_64 -d 0 -- -nographic -device e1000,netdev=net0,mac=50:05:00:02:00:00 -netdev tap,id=net0,ifname=vun003000500200,script=no ...

Continue reading the post on Patreon .