My Gateway picked up about 650Mb of traffic

Hi Folks,

Just checking, my little Dragino gateway with an external antenna seems to be picking up quite a lot of traffic since rebooted (and the antenna place 3 meters higher).

In the 17 days running it picked up about 650Mb of traffic and looking at the average payload size of each transmit/receive 650Mb seems a lot. Opening the console for just a few seconds shows numerous traffic, so I guess all is working fine.

Or am I misinterpreting the data?packet

Regards,

/rw

1 Like

How many uplinks & downlinks were in those 17 days?

Hi Nick,

Thanks for asking, if I open the Gateway console I’m seeing packets in/out, join requests flooding in like water in the Northsea.

/rw

Nice to hear it is working well.

Observing the traffic on a 25m high gateway I administer, I see a lot of “bogus” traffic unfortunately.
Very often, I see multiple join requests with the same APP ID and EUI id within a very short time.

Also I’ve seen endless join attempts with what looks like swapped APP ID and EUI id.
I think those can add up to quite a lot of traffic.

I’m not completely clear apologies… I’m seeing join requests/accepts but not as frequent as the regular traffic.

How many??? What were the counters over the 17 days?

Hi Nick, where can I find that?

/rw

found it:

Received Messages 340962

Transmitted Messages 14724

Using Bistromathematics:

650Mb over 17 days at 512 bytes per packet (add various TCP/IP headers and the housekeeping)

650,000,000 / 17 / 512 = 7,467

Certainly a few packets, but that’s 5 per minute. And if you are getting bad joins like @bertrik suggests, I do too, most of that is wasted repetitive requests to TTN which bumps up the count - although tracking down the device(s) may prove challenging.

1 Like

In a slightly varied Bistro,

17 days / 650MB / 340962 messages ==> 2kB per message (*)
20k messages/day
13 messages/minute

That s quite possible -
(my most visible gateway collects about 2k-10k messages per day)

*and there’s heartbeats & housekeeping which isn’t messages but creates traffic.

a tcpdump would seem to be a next useful step, to learn more.

ok, let me get some help on tcpdump, I got enough Linux gurus around me so that would work.

Funny (and fun) watching the traffic. In contrast to yours, my traffic dropped off pretty quickly at the end of October for no discernible reason. (I noticed it would plateau for a while after a reboot or update, but this seems to have fully flattened it.)

Screenshot 2020-11-10 at 09.56.08

(‘Fully flattened’ is a bit of an exaggeration - it’s just dropped to a <100 packets a day now.)

Hi Ryan,

How did you get the data to compile these graphs? Or are you making daily notes and put them in a sheet?

/rw

LoRa_Dern - your gateway naming goes to 11.

3 Likes

Heya. I have a Zabbix setup at home for monitoring various RPi-based services. I use a small Python script to grab the latest numbers from the gateway’s page every minute or so and pop them into the host’s dataset. It’s a terrible, no-good solution for a lot of reasons but it solved my immediate need of being able to broadly see if traffic was moving.