FR: Rate-limit error logs

[dseomn]

Hi,

My connection to my ISP recently went down for a short time, and dnss flooded the system logs with connect: network is unreachable errors. Would it be possible to add a rate limit to the logs so that, e.g., the same log message can't be logged more than once per second?

[albertito]

Hi! Thank you for opening this, sorry it took me so long to triage it.

This seems like a very reasonable feature request, will leave this open to track it :)

Do you have an example of the error messages you got overloaded with? So I know which ones to prioritize. No worries if not, I can simulate this if needed. Thank you!

[dseomn]

Here are the messages that were filling my logs. The first two had variations with different IP addresses. I'm slightly confused by the last two because dnss is running with --fallback_upstream=[::1]:14653. /etc/resolv.conf has nameserver 127.0.0.53, but I didn't think dnss used that.

server.go:105      resolver query error: POST failed: Post "https://dns.google/dns-query": dial tcp 8.8.4.4:443: connect: network is unreachable
server.go:106      dnsserver Handler: error: POST failed: Post \"https://dns.google/dns-query\": dial tcp 8.8.4.4:443: connect: network is unreachable
server.go:105      resolver query error: POST failed: Post "https://dns.google/dns-query": context deadline exceeded (Client.Timeout exceeded while awaiting headers)
server.go:106      dnsserver Handler: error: POST failed: Post \"https://dns.google/dns-query\": context deadline exceeded (Client.Timeout exceeded while awaiting headers)
server.go:105      resolver query error: POST failed: Post "https://dns.google/dns-query": dial tcp: lookup dns.google on 127.0.0.53:53: server misbehaving
server.go:106      dnsserver Handler: error: POST failed: Post \"https://dns.google/dns-query\": dial tcp: lookup dns.google on 127.0.0.53:53: server misbehaving