On Feb 11, 2022, at 8:33 AM, Tom Beecher <beec...@beecher.cc> wrote: > > The prediciate assumption that "pinging one destination is a valid check that > my internet works' is INCORRECT. There is no magical unicorn that could be > built that could make that true, and 'they're gonna do it anyways' is a poor > excuse to even consider it. >
The predicate assumption that unsuccessful pinging one destination is a valid check that my internet DOES NOT work' is ALSO INCORRECT. Still no magical unicorn. I am disappointed but not surprised to see this discussion on NANOG. Encouraging Users to use a tool (that is often ignored by the hardware targeted) by providing a non-revenue-creating special target does not make business sense. An allied issue is educating ‘Users’ about traceroute AKA sequential ping with TTL progression: Seeing missing or excessively long traceroute results from intermediate nodes does NOT indicate a real problem, especially when the target node is reachable with acceptable delay. I’ve lost count of my replies on user forums explaining this issue, even to otherwise well educated users. To be blunt, browsing to amazon.com, apple.com or another vendor site is a simple and easy to teach Internet aliveness check and, at least, offers the chance for the targeted vendor site to receive revenue from sales. I have no crisis of conscience from clicking an vendor shortcut for a basic end-to-end Internet functional test. Or for teaching a User to do the same. This meets the business purpose locally and requires no $pecial effort from Users, network providers, or target systems. This precludes memorization of IP addresses by end Users thus reducing the offered load from the likes of excessive ping 8.8.8.8. I would expect NANOG members to have favorite ping target addresses based on their environment, e.g., default router and a few designated targets. These are useful for manual debugging but, as mentioned previously, are not suitable as singular input to network monitoring.
