Dawnline 0.2.0
- Diagnostics page: Health and Log tabs in place of Health and Activity - Watchdog: repairs the AmneziaWG module, tunnels, nft table, firewall rules, uhttpd portal setting, portal, resolver and rpcd after two bad checks in a row, with backoff; after 3 failed repairs logs once in red, keeps repairing quietly and logs "restored" when it works again - Heartbeat cron job restarts a hung agent - Service log: tunnel and service starts with reasons (restarted, router start, update, crash), stops Dawnline made itself vs unexpected ones, certificate renewal and expiry, WAN address changes - Admin actions logged as "You ..." with their own icon and tint; every event has an icon tile; names kept for removed people and devices; server pill on events when there are several servers - Log filters: All, Connections, People, System, Problems; search and the command palette removed - Home network reaches people's devices: fw4 forwarding from the home zones, limited to people with home access - "Devices reach each other" per server, in the Access card - Portal: "This device" card first with a dotted border, key rotation and copy address in the menu, devices sorted Online/Offline/Not set up/Paused, over-limit slots in orange - People cards: slot bar, Online / Never connected, add chip always shown - Person and device pages: Add device tile, Connect at the bottom with File/Key/QR tabs, status text without dots; pulsing server tab dot - Server pill: Running / Not running / Turned off
This commit is contained in:
1 parent
d0955e44b8
commit
82ab86848e
50 files changed
+2166
-982
No files matched your search
Regular → Executable
+7
-3
@@ -1,5 +1,9 @@
|
||||
#!/bin/sh
|
||||
# Load devices into a Dawnline tunnel whenever netifd (re)creates it.
|
||||
[ "$ACTION" = ifup ] || exit 0
|
||||
# Dawnline tunnels: log starts and stops to the service log, and load devices
|
||||
# whenever netifd (re)creates one.
|
||||
case "$ACTION" in ifup|ifdown|ifup-failed) ;; *) exit 0 ;; esac
|
||||
[ -n "$(uci -q get "network.$INTERFACE._dawnline")" ] || exit 0
|
||||
/usr/bin/dawnline sync >/dev/null 2>&1 &
|
||||
# In order (a quick down/up must not log up first).
|
||||
/usr/bin/dawnline iface-event "$ACTION" "$INTERFACE" >/dev/null 2>&1
|
||||
[ "$ACTION" = ifup ] && /usr/bin/dawnline sync >/dev/null 2>&1 &
|
||||
exit 0
|
||||
@@ -55,6 +55,12 @@ start_service() {
|
||||
start_resolver
|
||||
}
|
||||
|
||||
# Noted for the service log: the next agent start is a restart, not a crash or a boot.
|
||||
stop_service() {
|
||||
mkdir -p /var/run/dawnline
|
||||
date +%s > /var/run/dawnline/agent.stopped
|
||||
}
|
||||
|
||||
reload_service() {
|
||||
/usr/bin/dawnline apply >/dev/null 2>&1
|
||||
start
|
||||
|
||||
@@ -24,6 +24,13 @@ rm -rf /usr/share/dawnline/portal/www/assets
|
||||
rm -f /tmp/luci-indexcache* /tmp/luci-modulecache/* 2>/dev/null
|
||||
touch /usr/lib/opkg/status 2>/dev/null
|
||||
|
||||
# Cron restarts the agent when it stops ticking (Health → Watchdog).
|
||||
mkdir -p /etc/crontabs
|
||||
grep -qs /usr/libexec/dawnline/heartbeat /etc/crontabs/root || {
|
||||
echo '*/5 * * * * /usr/libexec/dawnline/heartbeat' >> /etc/crontabs/root
|
||||
/etc/init.d/cron restart 2>/dev/null
|
||||
}
|
||||
|
||||
/etc/init.d/rpcd reload
|
||||
# A reload keeps LuCI logins but sometimes misses the plugin; restart only then.
|
||||
sleep 2
|
||||
|
||||
Reference in new issue
Block a user