Bug: [libxml2: getentropy failed with error code 38]

[bazirico]

Where did this bug occur? (required)

This bug occurs when I want to start the application.

• I am a self-hosted user reporting a bug from my self hosted app

Describe the bug

I'm running the Maybe containers on a Synology DS716+II.

I'm using ghcr.io/maybe-finance/maybe:stable but the same problem occurs with ghcr.io/maybe-finance/maybe:latest

Here's my docker-compose, I'm not sure you'll find relevant informations here :

version: "2.1"
networks:
  web:
    name: web
    external: true
services:
  maybe:
    image: ghcr.io/maybe-finance/maybe:stable
    labels:
      - "traefik.http.routers.maybe.rule=Host(`maybe.mysecretdomain.org`)"
      - "traefik.http.routers.maybe.service=maybe"
      - "traefik.http.routers.maybe.tls.certresolver=leresolver"
      - "traefik.http.routers.maybe.entrypoints=websecure"
      - "traefik.http.services.maybe.loadbalancer.server.port=3000"
    container_name: maybe
    environment:
      TZ: "Europe/Paris"
      SELF_HOSTED: "true"
      RAILS_FORCE_SSL: "false"
      RAILS_ASSUME_SSL: "false"
      GOOD_JOB_EXECUTION_MODE: async
      SECRET_KEY_BASE: ${SECRET_KEY_BASE:?}
      DB_HOST: postgres
      POSTGRES_DB: ${POSTGRES_DB:-maybe_production}
      POSTGRES_USER: ${POSTGRES_USER:-maybe_user}
      POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:?}
    volumes:
      - /volume1/docker/maybe/data/storage:/rails/storage
    networks:
      - web
    restart: unless-stopped
    depends_on:
      postgres:
        condition: service_healthy
  postgres:
    image: postgres:16
    restart: unless-stopped
    volumes:
      - /volume1/docker/maybe/data/postgres:/var/lib/postgresql/data
    environment:
      POSTGRES_USER: ${POSTGRES_USER:-maybe_user}
      POSTGRES_DB: ${POSTGRES_DB:-maybe_production}
      POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:?}
    healthcheck:
      test: [ "CMD-SHELL", "pg_isready -U $$POSTGRES_USER -d $$POSTGRES_DB" ]
      interval: 5s
      timeout: 5s
      retries: 5

When trying to start the application, the postgres container is running and healthy, but maybe container won't start :

bazirico@nas:/volume1/docker/maybe$ sudo docker-compose logs maybe
maybe  | libxml2: getentropy failed with error code 38
maybe  | /rails/bin/docker-entrypoint: line 6:     7 Aborted                 (core dumped) ./bin/rails db:prepare
maybe  | libxml2: getentropy failed with error code 38
maybe  | /rails/bin/docker-entrypoint: line 6:     7 Aborted                 (core dumped) ./bin/rails db:prepare
maybe  | libxml2: getentropy failed with error code 38
maybe  | /rails/bin/docker-entrypoint: line 6:     7 Aborted                 (core dumped) ./bin/rails db:prepare

To Reproduce
Steps to reproduce the behavior:

1. Follow the self hosted installation tutorial, then copy your docker-compose.yaml in your Synology NAS.
2. sudo docker-compose up -d
3. sudo docker-compose logs maybe
4. Check the error

Expected behavior
The application should start without any errors.

[zachgoll]

It looks like your custom network may be an issue here. I'm going to move this over to a discussion as this is not a bug with the core Maybe app; it's more of a setup configuration issue.

[bazirico]

Thank you for your feedback.

To be sure, I've tried with the docker-compose found in the tutorial here : https://raw.githubusercontent.com/maybe-finance/maybe/main/docker-compose.example.yml

Here's the output using this untouched docker-compose.yml :

bazirico@nas:/volume1/docker/maybe$ sudo docker-compose up
[+] Running 4/4
 ⠿ Volume "maybe_app-storage"    Created                                                                                                                                        0.2s
 ⠿ Volume "maybe_postgres-data"  Created                                                                                                                                        0.2s
 ⠿ Container maybe-postgres-1    Created                                                                                                                                        1.1s
 ⠿ Container maybe-app-1         Created                                                                                                                                        1.1s
Attaching to maybe-app-1, maybe-postgres-1
maybe-postgres-1  | The files belonging to this database system will be owned by user "postgres".
maybe-postgres-1  | This user must also own the server process.
maybe-postgres-1  |
maybe-postgres-1  | The database cluster will be initialized with locale "en_US.utf8".
maybe-postgres-1  | The default database encoding has accordingly been set to "UTF8".
maybe-postgres-1  | The default text search configuration will be set to "english".
maybe-postgres-1  |
maybe-postgres-1  | Data page checksums are disabled.
maybe-postgres-1  |
maybe-postgres-1  | fixing permissions on existing directory /var/lib/postgresql/data ... ok
maybe-postgres-1  | creating subdirectories ... ok
maybe-postgres-1  | selecting dynamic shared memory implementation ... posix
maybe-postgres-1  | selecting default max_connections ... 100
maybe-postgres-1  | selecting default shared_buffers ... 128MB
maybe-postgres-1  | selecting default time zone ... Etc/UTC
maybe-postgres-1  | creating configuration files ... ok
maybe-postgres-1  | running bootstrap script ... ok
maybe-postgres-1  | performing post-bootstrap initialization ... ok
maybe-postgres-1  | initdb: warning: enabling "trust" authentication for local connections
maybe-postgres-1  | initdb: hint: You can change this by editing pg_hba.conf or using the option -A, or --auth-local and --auth-host, the next time you run initdb.
maybe-postgres-1  | syncing data to disk ... ok
maybe-postgres-1  |
maybe-postgres-1  |
maybe-postgres-1  | Success. You can now start the database server using:
maybe-postgres-1  |
maybe-postgres-1  |     pg_ctl -D /var/lib/postgresql/data -l logfile start
maybe-postgres-1  |
maybe-postgres-1  | waiting for server to start....2025-02-24 14:05:17.196 UTC [55] LOG:  starting PostgreSQL 16.8 (Debian 16.8-1.pgdg120+1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 12.2.0-14) 12.2.0, 64-bit
maybe-postgres-1  | 2025-02-24 14:05:17.239 UTC [55] LOG:  listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
maybe-postgres-1  | 2025-02-24 14:05:17.385 UTC [58] LOG:  database system was shut down at 2025-02-24 14:05:14 UTC
maybe-postgres-1  | 2025-02-24 14:05:17.503 UTC [55] LOG:  database system is ready to accept connections
maybe-postgres-1  |  done
maybe-postgres-1  | server started
maybe-postgres-1  | CREATE DATABASE
maybe-postgres-1  |
maybe-postgres-1  |
maybe-postgres-1  | /usr/local/bin/docker-entrypoint.sh: ignoring /docker-entrypoint-initdb.d/*
maybe-postgres-1  |
maybe-postgres-1  | 2025-02-24 14:05:18.382 UTC [55] LOG:  received fast shutdown request
maybe-postgres-1  | waiting for server to shut down....2025-02-24 14:05:18.424 UTC [55] LOG:  aborting any active transactions
maybe-postgres-1  | 2025-02-24 14:05:18.431 UTC [55] LOG:  background worker "logical replication launcher" (PID 61) exited with exit code 1
maybe-postgres-1  | 2025-02-24 14:05:18.441 UTC [56] LOG:  shutting down
maybe-postgres-1  | 2025-02-24 14:05:18.510 UTC [56] LOG:  checkpoint starting: shutdown immediate
maybe-postgres-1  | .2025-02-24 14:05:20.091 UTC [56] LOG:  checkpoint complete: wrote 922 buffers (5.6%); 0 WAL file(s) added, 0 removed, 0 recycled; write=0.122 s, sync=0.955 s, total=1.650 s; sync files=301, longest=0.373 s, average=0.004 s; distance=4255 kB, estimate=4255 kB; lsn=0/1912110, redo lsn=0/1912110
maybe-postgres-1  | 2025-02-24 14:05:20.106 UTC [55] LOG:  database system is shut down
maybe-postgres-1  |  done
maybe-postgres-1  | server stopped
maybe-postgres-1  |
maybe-postgres-1  | PostgreSQL init process complete; ready for start up.
maybe-postgres-1  |
maybe-postgres-1  | 2025-02-24 14:05:20.457 UTC [1] LOG:  starting PostgreSQL 16.8 (Debian 16.8-1.pgdg120+1) on x86_64-pc-linux-gnu, compiled by gcc (Debian 12.2.0-14) 12.2.0, 64-bit
maybe-postgres-1  | 2025-02-24 14:05:20.480 UTC [1] LOG:  listening on IPv4 address "0.0.0.0", port 5432
maybe-postgres-1  | 2025-02-24 14:05:20.481 UTC [1] LOG:  listening on IPv6 address "::", port 5432
maybe-postgres-1  | 2025-02-24 14:05:20.560 UTC [1] LOG:  listening on Unix socket "/var/run/postgresql/.s.PGSQL.5432"
maybe-postgres-1  | 2025-02-24 14:05:20.646 UTC [71] LOG:  database system was shut down at 2025-02-24 14:05:20 UTC
maybe-postgres-1  | 2025-02-24 14:05:20.725 UTC [1] LOG:  database system is ready to accept connections
maybe-app-1       | libxml2: getentropy failed with error code 38
maybe-app-1       | /rails/bin/docker-entrypoint: line 6:     7 Aborted                 (core dumped) ./bin/rails db:prepare
maybe-app-1 exited with code 134
maybe-app-1       | libxml2: getentropy failed with error code 38
maybe-app-1       | /rails/bin/docker-entrypoint: line 6:     8 Aborted                 (core dumped) ./bin/rails db:prepare

I have no problem starting the app on my laptop, and this problem occurs only when I try to run Maybe on my Synology NAS (this NAS is already running my other personnal apps, approx. 20 containers).

Other people seem to have the same problem with getentropy() with particular libxml library versions and hardware combinaison, ex: SynoCommunity/spksrc#6439

Any thoughts ?

[bazirico]

It seems that even though I'm running the latest DSM version, my Synology NAS is still using an outdated kernel version:

bazirico@nas:~$ uname -a
Linux nas 3.10.108 #72806 SMP Thu Sep 5 13:39:20 CST 2024 x86_64 GNU/Linux synology_braswell_716+II

This is a limitation from Synology kernel politics. The function getentropy() was introduced in kernel 3.17, so my current setup does not support it.

🔗 Reference: GNU libc documentation on getentropy()

A potential solution mentioned here is to remove the use of getentropy().

Since I'm not a developer, I'm unsure about the implications of this change. Any insights or alternative solutions would be appreciated!

Also, could you move this back to the Issue section? I believe it will be helpful for other Synology users who might encounter the same problem.