Cloudron Forum

Hi all,

Since this morning my Contabo VPS (12 cores, 48GB RAM, 800GB NVMe) was very slow. I rebooted Contabo and now both n8n and Mautic won't start properly. The Contabo dashboard itself is also extremely slow.

Server specs: Contabo VPS 4 NVMe — 12 cores, 48GB RAM, 800GB NVMe
Resources are fine: 44GB RAM free, 663GB disk free

What happened after reboot:

Docker took ~5 minutes to start (stuck in "activating")
PostgreSQL took ~7 minutes for crash recovery (fsync of data directory), it wasn't shut down cleanly
n8n startup script got stuck on chown -R /app/data (2.3GB) for ~10 minutes due to slow disk
Once the chown finished, n8n enters a restart loop
Mautic also fails to start
n8n logs show:

Task runner connection attempt failed with status code 403 (repeated)
Task request timed out after 60 seconds
Database connection timed out / Database connection recovered (alternating)
Eventually receives SIGTERM and restarts
n8n never listens on port 5678
What I've verified:

PostgreSQL is up and accepting connections (pg_isready OK)
Can query the DB from inside the n8n container with psql — works fine
Port 5432 reachable, Redis running
All containers show as running in docker ps
n8n container: 0 restarts, using only 28MB RAM
n8n package: Updated Feb 9 (image 202602090744050000). Was working fine before the reboot.

Anyone know what I could do now?

Thanks!

Description

Our Cloudron environment is currently stuck during an infrastructure upgrade.
All installed apps are stuck showing:

• “Removing containers for upgrade”
• State: “Configuring”

This started when Cloudron attempted to upgrade the infrastructure from 49.8.0 → 49.9.0.
Since then, Cloudron cannot complete the upgrade and no apps can start anymore.

The system log indicates that Cloudron is unable to remove the Docker ‘cloudron’ network because Docker reports that the network still has active endpoints.
As a result the entire infrastructure upgrade aborts, leaving all apps stuck.

Steps to reproduce

1. Cloudron automatically initiated an infrastructure upgrade (49.8.0 → 49.9.0).
2. The platform stopped containers for the upgrade.
3. Cloudron attempted to remove and recreate the Docker ‘cloudron’ network.
4. Docker returned an error: network has active endpoints.
5. Infrastructure upgrade fails → apps remain in “configuring”.

Logs

Relevant portion from /home/yellowtent/platformdata/logs/box.log:

Nov 27 11:47:56 box:platform onActivated: starting post activation services
Nov 27 11:47:56 box:platform startInfra: checking infrastructure
Nov 27 11:47:56 box:platform markApps: reconfiguring apps
Nov 27 11:47:56 box:reverseproxy removeAppConfigs: removing app nginx configs
Nov 27 11:47:56 box:platform startInfra: updating infrastructure from 49.8.0 to 49.9.0
Nov 27 11:47:56 box:platform removeAllContainers: removing all containers for infra upgrade
Nov 27 11:47:56 box:shell platform: docker ps -qa --filter label=isCloudronManaged
Nov 27 11:47:57 box:platform createDockerNetwork: recreating docker network
Nov 27 11:47:57 box:shell platform: docker network rm -f cloudron
Nov 27 11:47:57 box:shell platform: docker network rm -f cloudron errored BoxError: docker exited with code 1 signal null
Nov 27 11:47:57 stderr: Error response from daemon: error while removing network cloudron: network has active endpoints
Nov 27 11:47:57 box:platform startInfra: Failed to start services. retry=false (attempt 0): docker exited with code 1 signal null

Troubleshooting Already Performed

• Rebooted the server
• Restarted Docker
• Restarted the Cloudron box service
• Verified that no non-Cloudron containers are running
• Apps remain stuck in “Removing containers for upgrade”

The issue persists every time Cloudron tries to start the infrastructure.

INFO
Cloudron version
9.0.12
Ubuntu version
Ubuntu 22.04.5 LTS Linux 5.15.0-161-generic
Vendor
QEMU
Product
Standard PC (i440FX + PIIX, 1996)
CPU
4 Core "AMD EPYC Processor (with IBPB)"
Memory
4.1 GB RAM & 4.12 GB Swap