Wrong RX time zone when GPS sync fails on relay gateway

3
Viewed 12

Hello,

I am using ChirpStack LNS with the LoRa protocol to retrieve data through a border gateway and a relay gateway.

My relay gateway uses GPS synchronization to set its system time. I noticed that when the gateway fails to synchronize with GPS (for example, when there is no GPS fix), the time field in the RX events is set to UTC+2, while my region is currently in UTC+1.

This results in the "time" value in the event up messages being incorrect by one hour.

While reviewing the concentrator configuration documentation (https://www.chirpstack.io/docs/chirpstack-concentratord/configuration.html
), I found the parameter:

time_fallback_enabled

I have set time_fallback_enabled = false, but I am not sure whether this actually resolves the issue or if it may introduce other side effects.

My questions are:

When GPS synchronization fails, how does ChirpStack determine the RX time value?

Does time_fallback_enabled = false ensure that no incorrect fallback time (e.g., local system time with wrong timezone) is used?

Is the RX time expected to always be in UTC regardless of the gateway’s local timezone?

What is the recommended configuration to avoid timezone inconsistencies when GPS is temporarily unavailable?

Thank you for your help.

1 Answers

I don't think this is a timezone issue, but an issue with the GNSS time synchronization related to the SX130X counter value (conversion between concentrator counter => UTC timestamp). It seems that when the GNSS fix is lost temporarily, this function starts returning incorrect timestamps. Last week I refactored the way this is implemented in the Concentratord, which should fix this issue. This is currently being tested.

Please see:
https://github.com/chirpstack/chirpstack-concentratord/commit/17c7d2c75349d4f0b0c7951f5140e4977a23a1a1