The logs were still flowing, but the terminal had locked up. You were running lnav. The culprit was a subtle Linux terminal bug that hits at the worst possible moment—when you need real-time insight.
This bug shows up when lnav interacts with certain terminal emulators or remote shells. The tool stops responding to key presses. Sometimes the output lags seconds behind. On some systems, it happens when a PTY buffer overfills or when TERM settings mismatch between local and remote environments. It’s easy to mistake for network latency, but the cause is deeper in the terminal layer.
Common triggers:
- Using lnav over SSH with tunneling through screen or tmux.
- Piping large log files into lnav instead of opening them directly.
- Systems with outdated ncurses libraries.
- Nonstandard TERM environment values (e.g.,
xterm-color vs xterm-256color).
Diagnosis steps:
- Run
echo $TERM and confirm it matches the local terminal capabilities. - Test lnav locally vs over SSH to isolate network factors.
- Use
script or strace to capture terminal activity during freeze events. - Update ncurses and verify lnav build flags.
Fixes that work:
- Upgrade lnav to the latest release; bugs tied to terminal handling have been patched in newer versions.
- Avoid piping huge datasets directly—load files into lnav from disk.
- Switch to a fully compatible terminal emulator like
xterm or kitty. - Set
TERM=xterm-256color before launching lnav to ensure correct escape handling.
For engineers dealing with log streaming on production boxes, this bug is more than annoying—it can block incident response. If your workflow depends on lnav, it’s worth scripting environment checks and preloading fixes into your session startup. That way the terminal is ready before your logs start screaming.
Stop waiting for patches to land upstream. Build a faster path. See how you can bypass the freeze and run lnav in a clean, reproducible environment—live in minutes—at hoop.dev.