Logs

Both openmixer units log to the systemd journal in the operator's user scope. There are no log files to rotate.

Reading them

journalctl --user -u openmixer-server -f            # follow the mixer
journalctl --user -u reac-pw -f              # follow the transport
journalctl --user -u openmixer-server -n 200 --no-pager
journalctl --user -u openmixer-server --since "-1h"
journalctl --user -u openmixer-server -b            # this boot only

From outside the operator's session — for example over SSH on a headless rig with lingering enabled:

machinectl shell <operator>@ /bin/journalctl --user -u openmixer-server -n 200 --no-pager

nginx logs the surface separately, in the system journal:

sudo journalctl -u nginx -n 100 --no-pager

What to collect when reporting a problem

{
  rpm -q openmixer openmixer-server openmixer-web-ui reac-pw libreac
  getcap /usr/bin/reac-pw
  systemctl --user status openmixer-server reac-pw
  ip -br link
  cat ~/.config/reac-pw/reac-pw.env ~/.config/reac-pw/*.env
  pw-metadata -n settings | grep clock
  curl -s http://127.0.0.1:8080/health
  curl -s http://127.0.0.1:8080/telemetry
  journalctl --user -u openmixer-server -n 300 --no-pager
  journalctl --user -u reac-pw  -n 300 --no-pager
} > /tmp/openmixer-report.txt 2>&1

That covers the questions that get asked first: which versions, whether the transport has its capabilities, whether the units are running, which interfaces exist, what the graph rate is, and what each process last said.

Turning up the transport's diagnostics

The transport is quiet by default. For a link that is misbehaving, enable its counters:

systemctl --user edit reac-pw

[Service]
Environment=REAC_DEBUG=1

systemctl --user restart reac-pw
journalctl --user -u reac-pw -f

Roughly every two seconds it then reports received / duplicate / other-source / bad frame counts and gap statistics, plus the ring's active channels, peak and fill.

Two of those numbers answer common questions directly:

duplicate greater than zero means frames are arriving twice — you are almost certainly on a mirrored switch port.
other-source counting up means more than one box is transmitting on the segment, and only the first one to establish is being decoded.

Turn it off again afterwards. It is diagnostic output, not something to leave running through a show.

In-mixer telemetry

The surface's telemetry panel reports what the logs cannot: the mains-path latency, the buffer quantum, the sample rate, a live xrun counter, and a per-channel, per-plugin latency breakdown. For dropout and latency questions, start there rather than in the journal.