Malformed Response on all relays since 22/09/16 5pm CEST

[logopk]

Hi, I'm facing a problem with resolution.

[2022-09-17 14:00:18] [ERROR] No valid network configuration for [odoh-cloudflare]
[2022-09-17 14:00:18] [NOTICE] dnscrypt-proxy is waiting for at least one server to be reachable
[2022-09-17 14:00:28] [NOTICE] Anonymizing queries for [odoh-crypto-sx] via [odohrelay-crypto-sx]
[2022-09-17 14:00:29] [WARNING] Unable to decrypt response from [odoh-crypto-sx]: [Malformed response]
[2022-09-17 14:00:29] [INFO] Trying to fetch the [odoh-crypto-sx] configuration again
[2022-09-17 14:00:29] [NOTICE] Anonymizing queries for [odoh-crypto-sx] via [odohrelay-crypto-sx]
[2022-09-17 14:00:31] [WARNING] Unable to decrypt response from [odoh-crypto-sx]: [Malformed response]

All of a sudden I had no DNS anymore. Fortunately I was able to connect my pihole to cloudflared (config from some time ago)

I have no idea how to troubleshoot. What info do you need - or maybe you immediately know the problem?! log_level does not get more output.

I can tell from the mtail logs (and my prometheus exporter) that it started with a NETWORK_ERROR and since then I get only NOT_READY response codes.

Thanks in advance

Peter

[ecliptik]

I had the same issue with the odohrelay-crypto-sx relay.

Switching to a different relay[1] and resolver[2] resolved the issue. I don't understand why odohrelay-crypto-sx suddenly started giving [Malformed response] message though.

1. https://raw.githubusercontent.com/DNSCrypt/dnscrypt-resolvers/master/v3/odoh-relays.md
2. https://raw.githubusercontent.com/DNSCrypt/dnscrypt-resolvers/master/v3/odoh-servers.md

[logopk]

How can I force dnscrypt-proxy to use another relay - the file is loaded from github and odohrelay-crypto-sx is #1...

[logopk]

How can I force dnscrypt-proxy to use another relay - the file is loaded from github and odohrelay-crypto-sx is #1...

routes = [
{ server_name='*', via=['odohrelay-crypto-sx', ''] }
]
????

[logopk]

routes = [ { server_name='*', via=['odohrelay-ibksturm', ''] } ] did the trick.

Is there a way to get this resolved automagically? Just in case ibksturm fails one day?

[ibksturm]

try this

routes = [ { server_name='*', via=['odohrelay-ibksturm', 'server2‘, ‘server3‘] } ]

watchup: i write with iphone, maybe the ‘ are the wring signens

[jedisct1]

This is due to a regression in zig when using the stage2 compiler.

A new version of the odoh-crypto-sx code compiled with the previous (stage1) compiler has been deployed, and the relay should be up and running again.

Thanks!