Help streaming Tomcat logs to standard out/err

We’re currently using this Lucee official image from Docker Hub (I think):


In production, we are running on Kubernetes (K8), which works well most of the time. But, sometimes, the K8 liveness probe (which hits port :8500 directly) just stops responding. And, nginx, on the same pod, can’t hit the upstream :8500 port either). This causes K8 to pull the pod out of rotation and restarted it.

I cannot find any evidence of why Lucee is becoming unresponsive. If I look in FusionReactor around the time of a given restart, all of the graphs look totally normal. Example (the missing data is the restart):

Just prior to the blip, all graphs look fine - no memory issues, no CPU issues, not requests pilling up. Smooth sailing. Then, out of no where, Lucee stops responding.

I saw in another thread that I might get a better idea if I look at the Tomcat logs (that it might be Tomcat refusing connections and not necessarily something with Lucee itself). In our platform, we just slurp everything from stderr and stdout into our log aggregator. So, if I can figure out how to pipe the Tomcat logs to the output, I can probably access them. But, this goes way beyond my day-to-day knowledge. I’ve tried Googling, but I can’t figure it out.

In the container, I can see that there is a file:


But, when I look at the contents, the last time-stamp is from 02-Nov-2020 (a year ago). So, I don’t think I’m even looking in the right place.

I’m totally lost :slight_smile: Any help would be much appreciated.

1 Like

I do see in the same folder that there is a date-stamped log file, catalina.2021-09-14.log, which has the right date. But, if I’m going to be doing something like symbolic-linking this file to /dev/stdout, then I probably can’t rely on knowing the date. :thinking:

does this help?

Sorry, I’m not very good with this low-level stuff. Are you saying that:

<system err="default" out="default"/>

in the lucee-server.xml will define how the Tomcat logs are written?