diff options
| author | Craig Jennings <c@cjennings.net> | 2026-07-14 00:23:16 -0500 |
|---|---|---|
| committer | Craig Jennings <c@cjennings.net> | 2026-07-14 00:23:16 -0500 |
| commit | aed151b1a49727a0faff15f25aba14e90ccc7a43 (patch) | |
| tree | 643b598ab70b8d613d59004c17139992dd9849ea /configs | |
| parent | 778a41118cb2c8b27677768dcf3890628746a8da (diff) | |
| download | archsetup-aed151b1a49727a0faff15f25aba14e90ccc7a43.tar.gz archsetup-aed151b1a49727a0faff15f25aba14e90ccc7a43.zip | |
feat(maint): ship a tailscale health-noise ignore list
The maint probe now reads network.tailscale_health_ignore before grading the tailscale lamp. This ships the fleet default: tailscaled's permanent DNS-unreachable complaint under systemd-resolved, which kept the lamp amber while DNS worked.
Diffstat (limited to 'configs')
| -rw-r--r-- | configs/maintenance-thresholds.toml | 7 |
1 files changed, 7 insertions, 0 deletions
diff --git a/configs/maintenance-thresholds.toml b/configs/maintenance-thresholds.toml index c14eaec..340c66c 100644 --- a/configs/maintenance-thresholds.toml +++ b/configs/maintenance-thresholds.toml @@ -102,6 +102,13 @@ unclean_boot_sample = 10 # boots assessed for the rate [network] ntp_offset_warn_ms = 100 # chrony last-offset bound (typical is sub-ms) +# Tailscale health messages that are standing noise on this fleet (substring +# match). tailscaled can't read resolved's stub/DoT config, so its +# DNS-unreachable complaint fires permanently while DNS works — it kept the +# lamp amber on every machine. A real DNS outage still surfaces elsewhere +# (clock sync, the net-tier scans). Ignored messages stay visible in the +# metric's evidence as health_ignored. +tailscale_health_ignore = ["can't reach the configured DNS servers"] [backups] rsyncshot_daily_crit_hours = 48 # daily backup older than this is a fire |