Gateways connectivity indicator shows offline (gray) in V2, no upgrading

Gateways
1

We are here in Asia, not planning to do an upgrade, just so that is clear. Since about 2 days my gateways are showing up as offline. Although messages come through. But it is quite a nuisance as the field staff relies on it in order to perform testing, and when it shows offline they assume the system is not working.

I feel that it has something to do with the following endpoint: http://noc.thethingsnetwork.org:8085/api/v2/gateways/
As you can see in the the timestamp there is 3 days old. I also use this in our dashboard as a source off online/offline status, is there another more reliable source for getting status messages?

{"timestamp":"2021-02-23T06:58:43.925595647Z","uplink":"22731","downlink":"2259","location":{},"frequency_plan":"AS_923_925","gps":{},"time":"1614063523925595647","rx_ok":22731,"tx_in":2259}
2

A quick search of the forum would tell you this is a know problem and one that pops up regularly with V2 Console, its not noc related directly rather noc scrapes the same flawed data source I believe. The fact that you can still see traffic routed through the gw’s to you applications show they are still working as expected, its just the console view that is wrong/not updating. Agree its frustrating esp if commissioning a new GW or running a ‘how to use TTN’ type training course (he says shedding a tear in frustration!), but all focus is now on proving and improving the new V3 infrastructure which is gradually rolling out through the rest of the year, and to which we will all be expected to migrate. It has been a particular problem since last weekend and there have been many forum posts on this in recent days, at one point >1/3rd of my own GW were showing off line :frowning: but fortunately most were still working and routing ok :slight_smile: unfortunately I knew a few were really down and getting through diagnosis without console assistance made trouble shooting harder :scream: … still have at least 4 showing offline (all showing ~3 days when I last checked) but which are routing traffic ok (when I checked this evening!). Best way to monitor if GW working ok is to deploy a nearby low power tx, short SF canary node - then you check RF reception, Packet handler/packet forwarder, compute platform and backhaul network connection all in one go :slight_smile: Look for traffic in an app and perhaps trigger sms or email if expected packet doesnt arrive, then you can investigate further…

4

Hi Jeff,

he says shedding a tear in frustration!

Classic :smiley:, oh the memories…

Thanks for the update, I think the sensor solution is a good solution, I can provision one such a unit and ship that out to the field, it is quite a hassle to get there (including a boat ride for 5 hours in a dinghy) but it is probably more reliable than the system we are having now.

The remoteness is also what makes me concerned about the whole upgrade to V3, but hey, I guess we see what happends, we first need to get south east asia on V3…

Cheers,

Roy

5

This sounds like a situation where on your next visit, you really, really, really want to deploy a remote access administration solution. When you do that, you also typically have the option of publishing the gateway’s impression of round-trip connectivity with TTN to some other data feed, eg, an MQTT broker of your preference.