add Watchdog disable before firmware upgrade (#414)

aredn · Sep 28, 2024 · 546ee21 · 546ee21
1 parent 956b308
commit 546ee21
Showing 1 changed file with 3 additions and 1 deletion.
diff --git a/arednGettingStarted/node_admin.rst b/arednGettingStarted/node_admin.rst
@@ -310,7 +310,9 @@ WAN web
   This switch enables http/https access to your node on its WAN interface. Disabling this option will not prevent http/https access to your node from the Mesh and LAN interfaces.
 
 Watchdog
-  Watchdog is a background monitor that keeps track of core node processes. If any of the processes has issues, Watchdog will reboot the node. This feature is ``disabled`` by default. Currently the set of node processes that are monitored include olsrd, dnsmasq, telnetd, dropbear, uhttpd, and vtund. Watchdog events are logged in the standard syslog on the node. Because the watchdog is part of the hardware, the device will still reboot itself even if the kernel crashes. If Watchdog is enabled, the following two fields will also be displayed.
+  Watchdog is a background monitor that keeps track of core node processes. If any of the processes has issues, Watchdog will reboot the node. This feature is ``disabled`` by default. Currently the set of node processes that are monitored include olsrd, dnsmasq, telnetd, dropbear, uhttpd, and vtund. Watchdog events are logged in the standard log on the node. Because the watchdog is part of the hardware, the device will still reboot itself even if the kernel crashes. Be aware that you must disable Watchdog and reboot your node before you can upgrade the firmware, since Watchdog may interfere with the normal upgrade process.
+
+  If Watchdog is enabled, the following two fields will also be displayed.
 
   Watchdog IP address
     You may also include one or more IP addresses, at least one of which should always be pingable. Your node will be rebooted if none of the IP addresses are reachable across the network. Enter IP addresses as a whitespace-delimited list. It is strongly recommended that you keep this list to the absolute minimum. Too many address can take a long time to check, especially if several are unavailable. This can result in reboots if the testing is not performed before the watchdog timer expires. Ideally use only one address.