bug: Continuwuity still queries DNS servers in /etc/resolv.conf even when dns_servers are present #2441

Open
opened 2026-11-02 17:25:49 +00:00 by stratself · 4 comments
Member

What is the bug?

Continuwuity still leaks DNS to resolvers inside /etc/resolv.conf, despite stuff being set inside dns_servers and accepted with !admin server show-config. I keep getting Matrix-related queries to my main DNS resolver, which ratelimits clients aggressively as it is meant for a different purpose. So this bug nullifies the benefits of the dns_servers variable itself.

Steps to reproduce

  1. Spin up two Technitium instances: one used for /etc/resolv.conf on host, and one configured to be included in c10y's dns_servers

  2. Watch dashboard/logs from both Technitiums. They both receive the starting netburst queries. Unsure about split but it's probably 50-50.

I've tested this on Podman in the following scenarios:

  • bridge network, where dns_servers point to another Technitium LAN IP on the same bridge network
  • pasta network mode, where c10y has a network_mode: container:continuwuity's-technitium and dns_servers = [ "127.0.0.1" ]
  • when mounting /etc/resolv.conf to the continuwuity container in both of the above cases, they talk to the nameserver inside that conf file. This is where I do not think it is a container issue anymore.

Server name and version

c10y main (ddbdd1e)

Additional context

As this still happens, I cannot rewrite docs to recommend using dns_servers.

### What is the bug? Continuwuity still leaks DNS to resolvers inside `/etc/resolv.conf`, despite stuff being set inside `dns_servers` and accepted with `!admin server show-config`. I keep getting Matrix-related queries to my main DNS resolver, which ratelimits clients aggressively as it is meant for a different purpose. So this bug nullifies the benefits of the `dns_servers` variable itself. ### Steps to reproduce 1. Spin up two Technitium instances: one used for `/etc/resolv.conf` on host, and one configured to be included in c10y's `dns_servers` 2. Watch dashboard/logs from both Technitiums. They both receive the starting netburst queries. Unsure about split but it's probably 50-50. I've tested this on Podman in the following scenarios: - bridge network, where `dns_servers` point to another Technitium LAN IP on the same bridge network - `pasta` network mode, where c10y has a `network_mode: container:continuwuity's-technitium` and `dns_servers = [ "127.0.0.1" ]` - when mounting `/etc/resolv.conf` to the continuwuity container in both of the above cases, they talk to the `nameserver` inside that conf file. This is where I do not think it is a container issue anymore. ### Server name and version c10y main (ddbdd1e) ### Additional context As this still happens, I cannot rewrite docs to recommend using `dns_servers`.
Author
Member
cc @silversys
Member

oh well thats just wonderful isnt it

ill take a look, thanks for the report

oh well thats just wonderful isnt it ill take a look, thanks for the report
Member

Question: What is the config value for query_all_nameservers? It defaults to true, unless you've changed it to false. I'm unsure if its the culprit as I'm still looking, but it would be an easy thing to check.

Question: What is the config value for `query_all_nameservers`? It defaults to true, unless you've changed it to false. I'm unsure if its the culprit as I'm still looking, but it would be an easy thing to check.
ginger changed title from bug: Continuwuity still queries the ones in /etc/resolv.conf even when dns_servers are present to bug: Continuwuity still queries DNS servers in /etc/resolv.conf even when dns_servers are present 2026-11-04 14:58:36 +00:00
Owner

@silversys wrote in #2441 (comment):

Question: What is the config value for query_all_nameservers? It defaults to true, unless you've changed it to false. I'm unsure if its the culprit as I'm still looking, but it would be an easy thing to check.

this was answered in the room as true (just to ensure it's documented)

https://matrix.to/#/!c10y-8Hm3mAW99j9KEKoWPQH2M_FxGwG-EVHK06iLKg/%24DqJbqdvZ9esnYbA9lJJ_WqGER12whcX56ZAOIfwqJAc?via=explodie.org&via=gingershaped.computer&via=federated.nexus

@silversys wrote in https://forgejo.ellis.link/continuwuation/continuwuity/issues/2441#issuecomment-38707: > Question: What is the config value for `query_all_nameservers`? It defaults to true, unless you've changed it to false. I'm unsure if its the culprit as I'm still looking, but it would be an easy thing to check. this was answered in the room as true (just to ensure it's documented) https://matrix.to/#/!c10y-8Hm3mAW99j9KEKoWPQH2M_FxGwG-EVHK06iLKg/%24DqJbqdvZ9esnYbA9lJJ_WqGER12whcX56ZAOIfwqJAc?via=explodie.org&via=gingershaped.computer&via=federated.nexus
Sign in to join this conversation.
No milestone
No assignees
3 participants
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set

Reference
continuwuation/continuwuity#2441
No description provided.