During adding SNTP (from the simple_SNTP sample in Zephyr) and later seen in other request system types, I've been getting a DNS recv error with error codes -4 and -103. This in nRF Connect SDK/Toolchain 2.7.0 on nrf7002dk rev1.0.2.
Logs example as shown below:
[00:00:15.118,560] <err> sntp: Failed to get time from SNTP server [00:00:15.128,540] <err> net_dns_resolve: DNS recv error (-4) [00:00:15.146,972] <err> net_dns_resolve: DNS recv error (-103)
This appears to be from the internal library, not from application code. The DNS request appears to later succeed - but I would like to either silence this first error or fix it.
Error code 4 is EINTR, and 103 does not appear to be a system error code but possibly could be either an invalid socket or an HTTP response of Early Hints (preparing the response), and if it is this HTTP error this shouldn't be considered an error.
On desktop OSes I know they can keep a local cache of DNS queries, or check the local network (I run a local DNS resolver on my network but these errors still show up on other regular networks) for a cached query - could this be the device checking for a cached query and failing?
The system checks to ensure a valid wifi connection prior to calling the sntp_init() function and still shows these errors even when given a heavy delay (have tried up to a minute). It usually ultimately succeeds to find a DNS entry.
Has anyone figured out how to fix this error or have it not be shown?
Thanks
