Hi everyone,
I’m trying to enable the AI Assistance / ExApp API on two self-hosted Nextcloud Hub 10 (31.0.7) servers and I’m hitting a wall during the Test-Deploy wizard.
Server 1 (GPU) | Server 2 (no GPU) | |
---|---|---|
OS | Ubuntu 24.04 LTS | Debian 12 Bookworm |
CPU / RAM / Disk | RTX 5090 × 1 16 GB RAM 512 GB NVMe |
16 GB RAM 512 GB SSD |
Nextcloud version | 31.0.7 | 31.0.7 |
LAN IP | 192.168.4.100 | 192.168.4.101 |
Both servers are on the same LAN; each runs its own Nextcloud instance.
- DSP container on Server 1
docker run -d --name nextcloud-appapi-dsp
–hostname nextcloud-appapi-dsp
–restart unless-stopped --privileged
-e NC_HAPROXY_PASSWORD=“NextCl0d2025”
-p 127.0.0.1:2375:2375
-v /var/run/docker.sock:/var/run/docker.sock
- Docker 23.0.6 +
nvidia-docker2
, driver 570 / CUDA 12.8 nvidia-smi
inside the container works.
- Daemon settings in NC-1 (same host as DSP)
Field (GUI) | Value |
---|---|
Daemon host | localhost:2375 |
Network | host |
Nextcloud URL | https://192.168.4.100 |
Additional opts | NC_SKIP_TLS_VERIFY=true |
Password | CHANGE_ME |
Compute device | CUDA (NVIDIA) |
- Check connection → success
- Test Deploy → stuck forever at “Init step” (
docker logs
shows:
ERROR occurred! Can't report the ExApp status to the Nextcloud instance.
)
- Daemon settings in NC-2 (on Server 2)
Field (GUI) | Value |
---|---|
Daemon host | http://192.168.4.100:2375 |
Network | host |
Nextcloud URL | https://192.168.4.101 |
Additional opts | NC_SKIP_TLS_VERIFY=true |
Password | same CHANGE_ME |
Compute device | CUDA (NVIDIA) |
Check connection ✓, but Test Deploy stalls at Heartbeat 250 (container stays healthy).
- What I tried / observations
5.1. From Server 1:
curl -sk -H “OCS-APIRequest: true”
https://192.168.4.100/ocs/v2.php/cloud/status
→ returns XML with <statuscode>998
Invalid query (HTTP 404).
5.2. Inside the test-deploy container:
Try default url to report the init status: https://192.168.4.100
5.3. * mod_rewrite
+ AllowOverride All
enabled; ran
occ maintenance:update:htaccess
— still 404.
- Certificates are self-signed, so I use
NC_SKIP_TLS_VERIFY=true
.
5.4. [Errno -2] Name or service not known
-
Questions
-
What usually causes Test Deploy to freeze at “Init step” with OCS 998 / 404?
-
On NC-2, why does the wizard pass health checks 1-249 then stop exactly at Heartbeat 250?
-
Is running one DSP for several Nextcloud instances on the LAN supported, or should I run a local DSP on each server?
-
What is the minimal Apache / Nextcloud setup (trusted domains, overwrite settings, rewrite rules) required for the wizard to finish?
-
Any additional AppAPI debug logs I can enable?
I’m still a beginner with Docker and server admin, so step-by-step hints or a sample working vHost would be awesome. Thanks in advance for any help!