Problem with active Public Apps in local deployed skipping proxy

[ghost]

Describe problem:
I'm deploying an application using docker compose with the manifest below. The deployment itself completes successfully. However, when I try to activate Public Apps, the system attempts to download configuration from the internet.
In our network environment, all outbound traffic from servers must go through a proxy to access external. The issue is that, based on the logs of shuffle-backend, it skips the proxy when trying to download the Public Apps configuration, which causes the download to fail:

2025/10/22 06:42:07 [DEBUG] Downloading API from URL https://shuffler.io/api/v1/apps/96a3d95a2a73cfdb51ea4a394287ed33/config

2025/10/22 06:42:07 [ERROR] Failed to split host and port: address shuffler.io: missing port in address

2025/10/22 06:42:07 [ERROR] Failed to split host and port for NOPROXY: address 10.19.135.245: missing port in address

2025/10/22 06:42:07 [INFO] Skipping proxy for https://shuffler.io/api/v1/apps/96a3d95a2a73cfdb51ea4a394287ed33/config

2025/10/22 06:42:09 [DEBUG] Looking for child workflows for workflow 181522b6-0042-451c-9710-cbe397e6b7ef to delete. User itsec-admin@mail.internal (39e23f70-c4bc-4be3-9c5b-8c4a4a4494f6) in org default (6d1a9df0-4e6a-4aa8-ac4f-2c5450c33349)

2025/10/22 06:42:09 [INFO] Should have deleted workflow Ops Dashboard Workflow (181522b6-0042-451c-9710-cbe397e6b7ef)

2025/10/22 06:42:09 [INFO] Deleted junk workflow with id: 181522b6-0042-451c-9710-cbe397e6b7ef

2025/10/22 06:42:09 [DEBUG] Platform health check successful! All necessary values are true.

2025/10/22 06:42:11 [INFO][dadb8202-a546-46fe-92cf-913ff77e25d7] Set workflowexecution based on new worker (>0.8.53) for workflow 181522b6-0042-451c-9710-cbe397e6b7ef. Actions: 45, Triggers: 0, Results: 45, Status: FINISHED

2025/10/22 06:42:37 [ERROR] Failed running auto-download request for 96a3d95a2a73cfdb51ea4a394287ed33: Get "https://shuffler.io/api/v1/apps/96a3d95a2a73cfdb51ea4a394287ed33/config": dial tcp 199.36.158.100:443: i/o timeout

My docker compose manifest:

opensearch-init:
  image: alpine
  container_name: opensearch-init
  volumes:
    - ./shuffle/shuffle-database:/usr/share/opensearch/data
  command: chown -R 1000:1000 /usr/share/opensearch/data
  restart: "no"

shuffle-frontend:
  image: ghcr.io/shuffle/shuffle-frontend:latest
  container_name: shuffle-frontend
  hostname: shuffle-frontend
  ports:
    - "${FRONTEND_PORT}:80"
    - "${FRONTEND_PORT_HTTPS}:443"
  networks:
    - cyber-blue
  environment:
    - BACKEND_HOSTNAME=${BACKEND_HOSTNAME}
  restart: unless-stopped
  depends_on:
    - shuffle-backend

shuffle-backend:
  image: ghcr.io/shuffle/shuffle-backend:latest
  container_name: shuffle-backend
  hostname: ${BACKEND_HOSTNAME}
  ports:
    - "${BACKEND_PORT}:5001"
  networks:
    - cyber-blue
  volumes:
    - /var/run/docker.sock:/var/run/docker.sock
    - ./shuffle/shuffle-apps:/shuffle-apps:z
    - ./shuffle/shuffle-files:/shuffle-files:z
  env_file: .env
  environment:
    - SHUFFLE_APP_HOTLOAD_FOLDER=/shuffle-apps
    - SHUFFLE_FILE_LOCATION=/shuffle-files
    - HTTP_PROXY=http://proxy-server.com:8080
    - HTTPS_PROXY=http://proxy-server.com:8080
    - NO_PROXY=10.19.135.245,localhost,${BACKEND_HOSTNAME},shuffle-frontend,shuffle-backend,${OUTER_HOSTNAME},shuffle-opensearch
  restart: unless-stopped

shuffle-orborus:
  image: ghcr.io/shuffle/shuffle-orborus:latest
  container_name: shuffle-orborus
  hostname: shuffle-orborus
  networks:
    - cyber-blue
  volumes:
    - /var/run/docker.sock:/var/run/docker.sock
  environment:
    - SHUFFLE_APP_SDK_TIMEOUT=300
    - SHUFFLE_ORBORUS_EXECUTION_CONCURRENCY=7
    - ENVIRONMENT_NAME=Shuffle
    - ORG_ID=Shuffle
    - BASE_URL=http://${OUTER_HOSTNAME}:5001
    - DOCKER_API_VERSION=1.40
    - SHUFFLE_WORKER_IMAGE=ghcr.io/shuffle/shuffle-worker:latest
    - SHUFFLE_STATS_DISABLED=true
    - SHUFFLE_LOGS_DISABLED=true
    - HTTP_PROXY=http://proxy-server.com:8080
    - HTTPS_PROXY=http://proxy-server.com:8080
    - NO_PROXY=10.19.135.245,localhost,${BACKEND_HOSTNAME},shuffle-frontend,shuffle-backend,${OUTER_HOSTNAME},shuffle-opensearch
  env_file: .env
  restart: unless-stopped
  security_opt:
    - seccomp:unconfined

shuffle-opensearch:
  image: opensearchproject/opensearch:3.0.0
  container_name: shuffle-opensearch
  hostname: shuffle-opensearch
  environment:
    - "OPENSEARCH_JAVA_OPTS=-Xms1024m -Xmx1024m"
    - bootstrap.memory_lock=true
    - DISABLE_PERFORMANCE_ANALYZER_AGENT_CLI=true
    - cluster.initial_master_nodes=shuffle-opensearch
    - cluster.routing.allocation.disk.threshold_enabled=false
    - cluster.name=shuffle-cluster
    - node.name=shuffle-opensearch
    - node.store.allow_mmap=false
    - discovery.seed_hosts=shuffle-opensearch
    - OPENSEARCH_INITIAL_ADMIN_PASSWORD=${SHUFFLE_OPENSEARCH_PASSWORD}
  ulimits:
    memlock:
      soft: -1
      hard: -1
    nofile:
      soft: 65536
      hard: 65536
  volumes:
    - ./shuffle/shuffle-database:/usr/share/opensearch/data:z
  ports:
    - 9201:9200
  networks:
    - cyber-blue
  restart: unless-stopped
  depends_on:
    - opensearch-init

Would really appreciate any insights, suggestions, or support to help solve this 🙏
Thanks.

[frikky]

Hey!

This is a mechanism for Shuffle to talk to our cloud platform shuffler.io to ensure it gets the latest app. If you however want the latest app, you can always manually import it!

Just go to https://shuffler.io/apps/ manually, or search for it. You can then download the specification at the top of the page and import it into your instance to create the app.

As for the proxy problem: I'll look into it, as it just looks like some kind of noproxy reading problem

[frikky]

Update: I found a noproxy parsing bug, and it will be a part of a new release soon.

[ghost]

Hi @frikky, thank you for answers 💯