Hey guys,
since a few weeks, I have the problem that I am not able to configure my Pangolin through docker labels. I'm giving the following labels to the frontend container of one of my applications, but I only get the message SendMessageInterval timed out after 10 attempts for message type: newt/wg/get-config for my newt container and nothing more happens.
labels:
- pangolin.public-resources.test-frontend.enabled=true
- pangolin.public-resources.test-frontend.name=test-frontend-site
- pangolin.public-resources.test-frontend.protocol=http
- pangolin.public-resources.test-frontend.full-domain=mydomain.com
- pangolin.public-resources.test-frontend.targets[0].hostname=frontend
- pangolin.public-resources.test-frontend.targets[0].port=3000
- pangolin.public-resources.test-frontend.targets[0].method=http
- pangolin.public-resources.test-frontend.auth.sso-enabled=true
- pangolin.public-resources.test-frontend.maintenance.type=automatic
- pangolin.public-resources.test-frontend.maintenance.title="We'll be back soon!"
- pangolin.public-resources.test-frontend.maintenance.message="We are currently performing scheduled maintenance. Please check back soon."
- pangolin.public-resources.test-frontend.maintenance.estimated-time="2 hours"
- pangolin.public-resources.test-frontend.rules[0].action=allow
- pangolin.public-resources.test-frontend.rules[0].match=ip
- pangolin.public-resources.test-frontend.rules[0].value=1.2.3.4
- pangolin.public-resources.test-frontend.rules[1].action=deny
- pangolin.public-resources.test-frontend.rules[1].match=cidr
- pangolin.public-resources.test-frontend.rules[1].value=0.0.0.0/0
For the Pangolin instance I get the following output
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Establishing websocket connection
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Client added to tracking - NEWT ID: XXX, Connection ID: XXX, Total connections: 1, Config version: 0
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: WebSocket connection fully established and ready - NEWT ID: XXX
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Handling ping request newt message!
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Sticking with previously connected node: Exit Node XXX (0 ms), latency diff = 0.0ms / NaN%.
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Public key mismatch. Deleting old peer...
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Deleting peer with public key XXX from exit node 1
gerbil-1 | INFO: 2026/04/08 08:38:12 Clearing connections for removed peer with WG IP: 10.0.0.20
gerbil-1 | INFO: 2026/04/08 08:38:12 Cleared 0 connections for WG IP: 10.0.0.20
gerbil-1 | INFO: 2026/04/08 08:38:12 Peer XXX removed successfully
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Adding peer with public key XXX to exit node 1
gerbil-1 | INFO: 2026/04/08 08:38:12 Clearing connections for added peer with WG IP: 10.0.0.20
gerbil-1 | INFO: 2026/04/08 08:38:12 Cleared 0 connections for WG IP: 10.0.0.20
gerbil-1 | INFO: 2026/04/08 08:38:12 Peer XXX added successfully
And in the web frontend, none of the options I gave with the labels is set.
The funny thing is that I remember that all was working fine at the beginning of the year and I haven't changed anything in my setup 🤔
Environment
OS Type & Version: (e.g., Ubuntu 22.04)
Debian and Ubuntu
Pangolin Version:
1.17.0
Gerbil Version:
1.3.1
Traefik Version:
latest container version
Newt Version:
1.11.0
To Reproduce
Setting up a test application with the docker labels set. Nothing happens in my pangolin site
Expected Behavior
The given options should be set in the Pangolin instance (which was working fine a few weeks ago)
Originally created by @daddeldash on GitHub (Apr 8, 2026).
Original GitHub issue: https://github.com/fosrl/pangolin/issues/2811
### Describe the Bug
Hey guys,
since a few weeks, I have the problem that I am not able to configure my Pangolin through docker labels. I'm giving the following labels to the frontend container of one of my applications, but I only get the message `SendMessageInterval timed out after 10 attempts for message type: newt/wg/get-config` for my newt container and nothing more happens.
```
labels:
- pangolin.public-resources.test-frontend.enabled=true
- pangolin.public-resources.test-frontend.name=test-frontend-site
- pangolin.public-resources.test-frontend.protocol=http
- pangolin.public-resources.test-frontend.full-domain=mydomain.com
- pangolin.public-resources.test-frontend.targets[0].hostname=frontend
- pangolin.public-resources.test-frontend.targets[0].port=3000
- pangolin.public-resources.test-frontend.targets[0].method=http
- pangolin.public-resources.test-frontend.auth.sso-enabled=true
- pangolin.public-resources.test-frontend.maintenance.type=automatic
- pangolin.public-resources.test-frontend.maintenance.title="We'll be back soon!"
- pangolin.public-resources.test-frontend.maintenance.message="We are currently performing scheduled maintenance. Please check back soon."
- pangolin.public-resources.test-frontend.maintenance.estimated-time="2 hours"
- pangolin.public-resources.test-frontend.rules[0].action=allow
- pangolin.public-resources.test-frontend.rules[0].match=ip
- pangolin.public-resources.test-frontend.rules[0].value=1.2.3.4
- pangolin.public-resources.test-frontend.rules[1].action=deny
- pangolin.public-resources.test-frontend.rules[1].match=cidr
- pangolin.public-resources.test-frontend.rules[1].value=0.0.0.0/0
```
Newt contaienr definition:
```
newt:
image: fosrl/newt
restart: unless-stopped
environment:
- PANGOLIN_ENDPOINT=https://pangolin.instance.tld
- NEWT_ID=${ID}
- NEWT_SECRET=${SECRET}
networks:
- frontend
volumes:
- /var/run/docker.sock:/var/run/docker.sock
```
For the Pangolin instance I get the following output
```
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Establishing websocket connection
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Client added to tracking - NEWT ID: XXX, Connection ID: XXX, Total connections: 1, Config version: 0
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: WebSocket connection fully established and ready - NEWT ID: XXX
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Handling ping request newt message!
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Sticking with previously connected node: Exit Node XXX (0 ms), latency diff = 0.0ms / NaN%.
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Public key mismatch. Deleting old peer...
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Deleting peer with public key XXX from exit node 1
gerbil-1 | INFO: 2026/04/08 08:38:12 Clearing connections for removed peer with WG IP: 10.0.0.20
gerbil-1 | INFO: 2026/04/08 08:38:12 Cleared 0 connections for WG IP: 10.0.0.20
gerbil-1 | INFO: 2026/04/08 08:38:12 Peer XXX removed successfully
pangolin-1 | 2026-04-08T08:38:12+00:00 [info]: Adding peer with public key XXX to exit node 1
gerbil-1 | INFO: 2026/04/08 08:38:12 Clearing connections for added peer with WG IP: 10.0.0.20
gerbil-1 | INFO: 2026/04/08 08:38:12 Cleared 0 connections for WG IP: 10.0.0.20
gerbil-1 | INFO: 2026/04/08 08:38:12 Peer XXX added successfully
```
And in the web frontend, none of the options I gave with the labels is set.
The funny thing is that I remember that all was working fine at the beginning of the year and I haven't changed anything in my setup 🤔
### Environment
- OS Type & Version: (e.g., Ubuntu 22.04)
- - Debian and Ubuntu
- Pangolin Version:
- - 1.17.0
- Gerbil Version:
- - 1.3.1
- Traefik Version:
- - latest container version
- Newt Version:
- 1.11.0
### To Reproduce
Setting up a test application with the docker labels set. Nothing happens in my pangolin site
### Expected Behavior
The given options should be set in the Pangolin instance (which was working fine a few weeks ago)
<!-- gh-comment-id:4210944285 -->
@oschwartz10612 commented on GitHub (Apr 9, 2026):
I think you need to enable the docker socket first.
https://docs.pangolin.net/manage/blueprints#enabling-docker-socket-access
I've already set a bind mount to the docker socket. And even the variable `DOCKER_SOCKET didn't change anything
<!-- gh-comment-id:4213367376 -->
@daddeldash commented on GitHub (Apr 9, 2026):
I've already set a bind mount to the docker socket. And even the variable `DOCKER_SOCKET didn't change anything
@LaurenceJJones commented on GitHub (Apr 9, 2026):
Can you set the Newt containter into debug mode and give us the output? typically it should print an error but maybe something happening?
just add environment
LOG_LEVEL=DEBUG
edit: cause I was testing blueprints yesterday and didnt have any issues.
<!-- gh-comment-id:4213393123 -->
@LaurenceJJones commented on GitHub (Apr 9, 2026):
Can you set the Newt containter into debug mode and give us the output? typically it should print an error but maybe something happening?
just add environment
```
LOG_LEVEL=DEBUG
```
_edit_: cause I was testing blueprints yesterday and didnt have any issues.
@LaurenceJJones commented on GitHub (Apr 13, 2026):
Okay so with Resource already exists, did you check the dashboard to see if the resource already exists?
<!-- gh-comment-id:4234924590 -->
@LaurenceJJones commented on GitHub (Apr 13, 2026):
Okay so with `Resource already exists`, did you check the dashboard to see if the resource already exists?
@LaurenceJJones commented on GitHub (Apr 13, 2026):
Within the logs its getting a target already for that site (so either: You delete target after newt startup, or the resource already existed), the labels are only checked when either:
Newt spins up
Container is created or destroyed (typically target application)
Could we take a step back and explain what your expecting to happen? as from the logs the resource and targets exist enough that this site connector is getting the configuration.
<!-- gh-comment-id:4235498125 -->
@LaurenceJJones commented on GitHub (Apr 13, 2026):
Within the logs its getting a target already for that site (so either: You delete target after newt startup, or the resource already existed), the labels are only checked when either:
- Newt spins up
- Container is created or destroyed (typically target application)
Could we take a step back and explain what your expecting to happen? as from the logs the resource and targets exist enough that this site connector is getting the configuration.
My expectation is that I run the containers and pangolin configures everything like defined with the labels. My goal is to use pangolin without touching it by hand, what is significantly facilitated by the provisioning keys (thanks a lot for this!!!).
I had another try, where I deleted the site and all related resources, turned on the debug mode for newt and the result was that I now have a site without any resources.
@LaurenceJJones commented on GitHub (Apr 13, 2026):
My expectation is that I run the containers and pangolin configures everything like defined with the labels. My goal is to use pangolin without touching it by hand, what is significantly facilitated by the provisioning keys (thanks a lot for this!!!).
I had another try, where I deleted the site and all related resources, turned on the debug mode for newt and the result was that I now have a site without any resources.
That because the docker socket environment or flag is not set
The key line is:
DEBUG: 2026/04/13 10:53:08 Docker socket path is not set
So this looks like a Docker socket configuration issue (missing --docker-socket / DOCKER_SOCKET). Did you ensure that this exists as per the earlier conversation?
<!-- gh-comment-id:4235957029 -->
@LaurenceJJones commented on GitHub (Apr 13, 2026):
> My expectation is that I run the containers and pangolin configures everything like defined with the labels. My goal is to use pangolin without touching it by hand, what is significantly facilitated by the provisioning keys (thanks a lot for this!!!).
>
> I had another try, where I deleted the site and all related resources, turned on the debug mode for newt and the result was that I now have a site without any resources.
>
> ```
> newt-1 | INFO: 2026/04/13 10:53:07 Newt version 1.11.0
> newt-1 | DEBUG: 2026/04/13 10:53:07 Starting metrics server on 127.0.0.1:2112
> newt-1 | INFO: 2026/04/13 10:53:08 Loading config from: /root/.config/newt-client/config.json
> newt-1 | INFO: 2026/04/13 10:53:08 Config file does not exist at /root/.config/newt-client/config.json, will create it with provided values
> newt-1 | DEBUG: 2026/04/13 10:53:08 Endpoint: https://pangolin.instance.tld
> newt-1 | DEBUG: 2026/04/13 10:53:08 Log Level: DEBUG
> newt-1 | DEBUG: 2026/04/13 10:53:08 Docker Network Validation Enabled: false
> newt-1 | DEBUG: 2026/04/13 10:53:08 Health Check Certificate Enforcement: false
> newt-1 | DEBUG: 2026/04/13 10:53:08 Dns: 9.9.9.9
> newt-1 | DEBUG: 2026/04/13 10:53:08 MTU: 1280
> newt-1 | DEBUG: 2026/04/13 10:53:08 Setting up clients with netstack2...
> newt-1 | DEBUG: 2026/04/13 10:53:08 Created shared UDP socket on port 51392 (refcount: 2)
> newt-1 | DEBUG: 2026/04/13 10:53:08 Creating new health check monitor with certificate enforcement: false
> newt-1 | INFO: 2026/04/13 10:53:08 Provisioning key found – exchanging for newt credentials...
> newt-1 | DEBUG: 2026/04/13 10:53:08 Provisioning response body: {"data":{"newtId":"XXXXXXXXXX","secret":"XXXXXXXXXX"},"success":true,"error":false,"message":"Newt registered successfully","status":201}
> newt-1 | INFO: 2026/04/13 10:53:08 Successfully provisioned – newt ID: XXXXXXXXXX
> newt-1 | INFO: 2026/04/13 10:53:08 Saving config to: /root/.config/newt-client/config.json
> newt-1 | DEBUG: 2026/04/13 10:53:08 Token response body: {"data":{"token":"XXXXXXXXXX","serverVersion":"1.17.0"},"success":true,"error":false,"message":"Token created successfully","status":200}
> newt-1 | INFO: 2026/04/13 10:53:08 Server version: 1.17.0
> newt-1 | DEBUG: 2026/04/13 10:53:08 Received token: XXXXXXXXXX
> newt-1 | DEBUG: 2026/04/13 10:53:08 Config has not changed, skipping save
> newt-1 | DEBUG: 2026/04/13 10:53:08 Public key: XXXXXXXXXX
> newt-1 | INFO: 2026/04/13 10:53:08 Websocket connected
> newt-1 | DEBUG: 2026/04/13 10:53:08 Requesting exit nodes from server
> newt-1 | DEBUG: 2026/04/13 10:53:08 Requesting WireGuard configuration from remote server
> newt-1 | INFO: 2026/04/13 10:53:08 Provisioning detected – sending provisioning blueprint
> newt-1 | DEBUG: 2026/04/13 10:53:08 Sending message: newt/wg/register, data: map[backwardsCompatible:true chainId:ecaec72c97e67109 newtVersion:1.11.0 publicKey:XXXXXXXXXX]
> newt-1 | DEBUG: 2026/04/13 10:53:08 Sending message: newt/ping/request, data: map[chainId:6d2e36062869ce87 noCloud:false]
> newt-1 | DEBUG: 2026/04/13 10:53:08 Sending message: newt/wg/get-config, data: map[chainId:e0b35b5b0e2d5e04 port:51392 publicKey:XXXXXXXXXX]
> newt-1 | DEBUG: 2026/04/13 10:53:08 Received ping message
> newt-1 | DEBUG: 2026/04/13 10:53:08 Only one exit node available, using it directly: pangolin.instance.tld
> newt-1 | DEBUG: 2026/04/13 10:53:08 Sending message: newt/wg/register, data: map[chainId:fb5822618a03d581 newtVersion:1.11.0 pingResults:[{ExitNodeID:1 LatencyMs:0 Weight:1 Error: Name:Exit Node 0Z8SWFfG Endpoint:pangolin.instance.tld WasPreviouslyConnected:false}] publicKey:XXXXXXXXXX]
> newt-1 | DEBUG: 2026/04/13 10:53:08 Received Docker container fetch request
> newt-1 | DEBUG: 2026/04/13 10:53:08 Docker socket path is not set
> newt-1 | DEBUG: 2026/04/13 10:53:08 Received registration message
> newt-1 | DEBUG: 2026/04/13 10:53:08 Received registration message data: map[chainId:fb5822618a03d581 endpoint:pangolin.instance.tld:51821 healthCheckTargets:[] publicKey:XXXXXXXXXX relayPort:21820 serverIP:10.0.0.1 targets:map[tcp:[] udp:[]] tunnelIP:10.0.0.60]
> newt-1 | DEBUG: 2026/04/13 10:53:08 Received: {Type:newt/wg/connect Data:map[chainId:fb5822618a03d581 endpoint:pangolin.instance.tld:51821 healthCheckTargets:[] publicKey:XXXXXXXXXX relayPort:21820 serverIP:10.0.0.1 targets:map[tcp:[] udp:[]] tunnelIP:10.0.0.60] ConfigVersion:0}
> newt-1 | INFO: 2026/04/13 10:53:08 Connecting to endpoint: pangolin.instance.tld
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: decryption worker 1 - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: handshake worker 1 - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: encryption worker 2 - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: decryption worker 2 - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: handshake worker 2 - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: TUN reader - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: event worker - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Interface up requested
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 UDP bind has been updated
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Interface state was Down, requested Up, now Up
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: receive incoming v6 - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: encryption worker 1 - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 Routine: receive incoming v4 - started
> newt-1 | DEBUG: 2026/04/13 10:53:08 Starting UDP hole punch to 1 exit nodes with shared bind
> newt-1 | DEBUG: 2026/04/13 10:53:08 Starting hole punch to 1.2.3.4 with public key: XXXXXXXXXX
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 UAPI: Updating private key
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - UAPI: Created
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - UAPI: Adding allowedip
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - UAPI: Updating endpoint
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - UAPI: Updating persistent keepalive interval
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - Starting
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - Sending keepalive packet
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - Sending handshake initiation
> newt-1 | DEBUG: 2026/04/13 10:53:08 WireGuard device created. Lets ping the server now...
> newt-1 | DEBUG: 2026/04/13 10:53:08 Testing initial connection with reliable ping...
> newt-1 | DEBUG: 2026/04/13 10:53:08 Resolved exit node: 1.2.3.4 -> 1.2.3.4:21820
> newt-1 | DEBUG: 2026/04/13 10:53:08 Sent UDP hole punch to 1.2.3.4:21820: {"ephemeralPublicKey":"XXXXXXXXXX","nonce":"Z9EFSTk26TQxx+Qv","ciphertext":"XXXXXXXXXX"}
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - Routine: sequential sender - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - Routine: sequential receiver - started
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:08 peer(0Z8S…o4k4) - Received handshake response
> newt-1 | DEBUG: 2026/04/13 10:53:08 Initial connection test successful
> newt-1 | DEBUG: 2026/04/13 10:53:08 Ping attempt 1
> newt-1 | DEBUG: 2026/04/13 10:53:08 Ping latency: 18.018362ms
> newt-1 | INFO: 2026/04/13 10:53:08 Tunnel connection to server established successfully!
> newt-1 | DEBUG: 2026/04/13 10:53:08 Starting ping check
> newt-1 | DEBUG: 2026/04/13 10:53:08 Started direct UDP relay on 10.0.0.60:51392 (bidirectional via SharedBind)
> newt-1 | DEBUG: 2026/04/13 10:53:08 Adding 0 health check targets in bulk
> newt-1 | DEBUG: 2026/04/13 10:53:08 Successfully added all 0 health check targets
> newt-1 | DEBUG: 2026/04/13 10:53:08 Successfully added 0 health check targets
> newt-1 | DEBUG: 2026/04/13 10:53:08 Direct UDP relay started (bidirectional through SharedBind)
> newt-1 | DEBUG: 2026/04/13 10:53:09 Sent UDP hole punch to 1.2.3.4:21820: {"ephemeralPublicKey":"XXXXXXXXXX"}
> newt-1 | DEBUG: 2026/04/13 10:53:09 Increased hole punch interval to 2s
> newt-1 | DEBUG: 2026/04/13 10:53:10 Sending message: newt/wg/get-config, data: map[chainId:e0b35b5b0e2d5e04 port:51392 publicKey:XXXXXXXXXX]
> newt-1 | DEBUG: 2026/04/13 10:53:11 Sent UDP hole punch to 1.2.3.4:21820: {"ephemeralPublicKey":"XXXXXXXXXX","nonce":"mWEi/geDCgxWq4uN","ciphertext":"XXXXXXXXXX"}
> newt-1 | DEBUG: 2026/04/13 10:53:11 Increased hole punch interval to 4s
> newt-1 | DEBUG: 2026/04/13 10:53:12 Sending message: newt/wg/get-config, data: map[chainId:e0b35b5b0e2d5e04 port:51392 publicKey:XXXXXXXXXX]
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:13 peer(0Z8S…o4k4) - Sending keepalive packet
> newt-1 | DEBUG: 2026/04/13 10:53:14 Sending message: newt/wg/get-config, data: map[chainId:e0b35b5b0e2d5e04 port:51392 publicKey:XXXXXXXXXX]
> newt-1 | DEBUG: 2026/04/13 10:53:15 Sent UDP hole punch to 1.2.3.4:21820: {"ephemeralPublicKey":"XXXXXXXXXX","nonce":"VmNj+SAm0GvzOu+6","ciphertext":"XXXXXXXXXX"}
> newt-1 | DEBUG: 2026/04/13 10:53:15 Increased hole punch interval to 8s
> newt-1 | DEBUG: 2026/04/13 10:53:16 Sending message: newt/wg/get-config, data: map[chainId:e0b35b5b0e2d5e04 port:51392 publicKey:XXXXXXXXXX]
> newt-1 | DEBUG: 2026/04/13 10:53:18 Sending message: newt/wg/get-config, data: map[chainId:e0b35b5b0e2d5e04 port:51392 publicKey:XXXXXXXXXX]
> newt-1 | DEBUG: gerbil-wireguard: 2026/04/13 10:53:18 peer(0Z8S…o4k4) - Sending keepalive packet
> newt-1 | DEBUG: 2026/04/13 10:53:20 Sending message: newt/wg/get-config, data: map[chainId:e0b35b5b0e2d5e04 port:51392 publicKey:XXXXXXXXXX]
> ```
That because the docker socket environment or flag is not set
The key line is:
DEBUG: 2026/04/13 10:53:08 Docker socket path is not set
So this looks like a Docker socket configuration issue (missing --docker-socket / DOCKER_SOCKET). Did you ensure that this exists as per the earlier conversation?
Is this something that you changed on some point? I've never set the DOCKER_SOCKET variable and it always worked. Now I have to set the variable explicitly to enable the communication with docker? If I do, it works, but wouldn't it be better to test if newt can talk to docker.sock and if yes, enable it with the default docker.sock path?
<!-- gh-comment-id:4235985216 -->
@daddeldash commented on GitHub (Apr 13, 2026):
Is this something that you changed on some point? I've never set the DOCKER_SOCKET variable and it always worked. Now I have to set the variable explicitly to enable the communication with docker? If I do, it works, but wouldn't it be better to test if newt can talk to docker.sock and if yes, enable it with the default docker.sock path?
@LaurenceJJones commented on GitHub (Apr 13, 2026):
Is this something that you changed on some point? I've never set the DOCKER_SOCKET variable and it always worked. Now I have to set the variable explicitly to enable the communication with docker? If I do, it works, but wouldn't it be better to test if newt can talk to docker.sock and if yes, enable it with the default docker.sock path?
Yes it was changed to opt in rather than defaulting to reading everything from a security point of view we shouldnt guess a user wants docker integration simply if the socket exists, as when deploying on bare host the socket may exist but not to be used.
<!-- gh-comment-id:4236010102 -->
@LaurenceJJones commented on GitHub (Apr 13, 2026):
> Is this something that you changed on some point? I've never set the DOCKER_SOCKET variable and it always worked. Now I have to set the variable explicitly to enable the communication with docker? If I do, it works, but wouldn't it be better to test if newt can talk to docker.sock and if yes, enable it with the default docker.sock path?
Yes it was changed to [opt in](https://github.com/fosrl/newt/commit/eb8a12f2906e0087501d73db2a2262d89710972d) rather than defaulting to reading everything from a security point of view we shouldnt guess a user wants docker integration simply if the socket exists, as when deploying on bare host the socket may exist but not to be used.
Ah, that explains almost everything 😅 Thanks for your support!!! I really appreciate that
<!-- gh-comment-id:4236019966 -->
@daddeldash commented on GitHub (Apr 13, 2026):
Ah, that explains almost everything 😅 Thanks for your support!!! I really appreciate that
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
Originally created by @daddeldash on GitHub (Apr 8, 2026).
Original GitHub issue: https://github.com/fosrl/pangolin/issues/2811
Describe the Bug
Hey guys,
since a few weeks, I have the problem that I am not able to configure my Pangolin through docker labels. I'm giving the following labels to the frontend container of one of my applications, but I only get the message
SendMessageInterval timed out after 10 attempts for message type: newt/wg/get-configfor my newt container and nothing more happens.Newt contaienr definition:
For the Pangolin instance I get the following output
And in the web frontend, none of the options I gave with the labels is set.
The funny thing is that I remember that all was working fine at the beginning of the year and I haven't changed anything in my setup 🤔
Environment
To Reproduce
Setting up a test application with the docker labels set. Nothing happens in my pangolin site
Expected Behavior
The given options should be set in the Pangolin instance (which was working fine a few weeks ago)
@oschwartz10612 commented on GitHub (Apr 9, 2026):
I think you need to enable the docker socket first.
https://docs.pangolin.net/manage/blueprints#enabling-docker-socket-access
@daddeldash commented on GitHub (Apr 9, 2026):
I've already set a bind mount to the docker socket. And even the variable `DOCKER_SOCKET didn't change anything
@LaurenceJJones commented on GitHub (Apr 9, 2026):
Can you set the Newt containter into debug mode and give us the output? typically it should print an error but maybe something happening?
just add environment
edit: cause I was testing blueprints yesterday and didnt have any issues.
@daddeldash commented on GitHub (Apr 9, 2026):
It tells me somewhere in the log, that it is failing with
Resource already exists. But this is new to me 🤔Anyway...here is the debug output
@daddeldash commented on GitHub (Apr 13, 2026):
Why is this ticket closed? 🤔
@LaurenceJJones commented on GitHub (Apr 13, 2026):
Okay so with
Resource already exists, did you check the dashboard to see if the resource already exists?@daddeldash commented on GitHub (Apr 13, 2026):
I've removed the resource and restartet the docker containers. This is the logfile
@daddeldash commented on GitHub (Apr 13, 2026):
The resource is still not created
@LaurenceJJones commented on GitHub (Apr 13, 2026):
Within the logs its getting a target already for that site (so either: You delete target after newt startup, or the resource already existed), the labels are only checked when either:
Could we take a step back and explain what your expecting to happen? as from the logs the resource and targets exist enough that this site connector is getting the configuration.
@daddeldash commented on GitHub (Apr 13, 2026):
My expectation is that I run the containers and pangolin configures everything like defined with the labels. My goal is to use pangolin without touching it by hand, what is significantly facilitated by the provisioning keys (thanks a lot for this!!!).
I had another try, where I deleted the site and all related resources, turned on the debug mode for newt and the result was that I now have a site without any resources.
@LaurenceJJones commented on GitHub (Apr 13, 2026):
That because the docker socket environment or flag is not set
The key line is:
DEBUG: 2026/04/13 10:53:08 Docker socket path is not set
So this looks like a Docker socket configuration issue (missing --docker-socket / DOCKER_SOCKET). Did you ensure that this exists as per the earlier conversation?
@daddeldash commented on GitHub (Apr 13, 2026):
Is this something that you changed on some point? I've never set the DOCKER_SOCKET variable and it always worked. Now I have to set the variable explicitly to enable the communication with docker? If I do, it works, but wouldn't it be better to test if newt can talk to docker.sock and if yes, enable it with the default docker.sock path?
@LaurenceJJones commented on GitHub (Apr 13, 2026):
Yes it was changed to opt in rather than defaulting to reading everything from a security point of view we shouldnt guess a user wants docker integration simply if the socket exists, as when deploying on bare host the socket may exist but not to be used.
@daddeldash commented on GitHub (Apr 13, 2026):
Ah, that explains almost everything 😅 Thanks for your support!!! I really appreciate that