Role of NTP on RAK7268V2s

Hello,

I have a large number of RAK7268V2s deployed at various locations running basic station mode with AWS IoT Core as the LNS. One issue I have run into is that some of these locations block NTP traffic and this seems to cause the gateways to be unable to transmit data to AWS. I know that clock drift can cause various issues to any application, but I have not found any documentation around how NTP is used by these gateways. My theory is that that the gateways when the gateways are unable to maintain their clock they will operate until the clock drifts too far out of sync. It seems that sometimes the hubs are able to maintain their clock through a power interruption and other times they will lose their clock immediately. My guess is that the gateways have a CMOS battery and some of the gateways have depleted the power in this battery. I have limited visibility to these gateways, so this is all speculation from various testing scenarios. Is there anyone who has dealt with these kinds of issues or can share some information that may be helpful?

Thank you,