A cross platform golang program to check for Internet outages via DNS for devices behind a AT&T Fiber gateway.
- Detects outages
- Can generate and report outage metric to statsd(Datadog)
- Custom Datadog dashboards using the metric
- Custom Datadog monitors using the metric for alerting
- Can be run as a daemon in kubernetes via helm, docker, or systemd
It works most reliably on a device in the gateway's subnet, the default being 192.168.1.0/24. This means when using passthrough you need a device outside the passthrough, and not the passthrough device. This was observed with time, and the exact cause is unknown.
- BGW320-505 gateway
- BGW320-500 gateway, expected to work, but untested
- BGW210-700 gateway, expected to work, but untested
By default the AT&T Fiber gateway intercepts A record DNS queries during an outage. It returns the ipv4 address of the gateway as the answer. So if the ipv4 address returned matches the ipv4 address of the gateway an outage has been detected. By default it loops to keep detecting outages.
Example:
nslookup google.com 8.8.8.8
Server: 8.8.8.8
Address: 8.8.8.8#53
Non-authoritative answer:
Name: google.com
Address: 192.168.1.254
Here is the documentation.
I have been experiencing intermittent outages with my AT&T Fiber. After having
AT&T upgrade my gateway to firmware version 6.30.5 it has been more stable.
Especially with the issue of the gateway rebooting itself. Though that still
isn't perfect.
Yet I continue to see at least one 1-2 minute outage daily. I have monitoring the network traffic on the outside interface of my personal router, setup in passthrough mode, with tcpdump. During an outage I could clearly see the gateway intercepting DNS queries for A records.
I had been testing a script and program to detect outages by detecting it intercepting web traffic. This worked, but was more error prone and complex. This method is much simpler, and should always work.
Here is a link to another file with examples of the output from tcpdump showing the interception during an outage.