Container won't start.

[giangitesta]

Hi @chvvkumar,
i tried to create the container as per the instructions, but unfortunately it won't start.

It seems to have problems starting the Alpaca SafetyMonitor service (Waitress + MQTT unified service), and after a few seconds it exits (with code 132).

Permissions throughout the application tree are set to 777 to avoid problems.
Image

What do you recommend I do?
I don't see much log to analyse.

Best regards
Gianluca

[chvvkumar]

Hi!

Normally, exit code 132 indicates mismatch of architecture type between the container and the docker host. Can you post where you are running the container and also any logs from docker logs -f simple-cloud-detect?

[giangitesta]

Hi,
this is the log after the execution:

gianluca@athena:~$ docker logs -f simple-cloud-detect

Starting Unified Cloud Detection Service...
Converting model if needed...
2025-12-28 16:37:26,027 - INFO - No 'groups' attribute found in model config
Starting Alpaca SafetyMonitor (Waitress + MQTT unified service)...

and the system information:

gianluca@athena:~$ docker info
Client: Docker Engine - Community
Version: 28.5.1
Context: default
Debug Mode: false
Plugins:
buildx: Docker Buildx (Docker Inc.)
Version: v0.29.1
Path: /usr/libexec/docker/cli-plugins/docker-buildx
compose: Docker Compose (Docker Inc.)
Version: v2.40.2
Path: /usr/libexec/docker/cli-plugins/docker-compose

Server:
Containers: 17
Running: 15
Paused: 0
Stopped: 2
Images: 15
Server Version: 28.5.1
Storage Driver: overlay2
Backing Filesystem: extfs
Supports d_type: true
Using metacopy: false
Native Overlay Diff: true
userxattr: false
Logging Driver: json-file
Cgroup Driver: systemd
Cgroup Version: 2
Plugins:
Volume: local
Network: bridge host ipvlan macvlan null overlay
Log: awslogs fluentd gcplogs gelf journald json-file local splunk syslog
CDI spec directories:
/etc/cdi
/var/run/cdi
Swarm: inactive
Runtimes: io.containerd.runc.v2 runc
Default Runtime: runc
Init Binary: docker-init
containerd version: b98a3aace656320842a23f4a392a33f46af97866
runc version: v1.3.0-0-g4ca628d1
init version: de40ad0
Security Options:
apparmor
seccomp
Profile: builtin
cgroupns
Kernel Version: 6.8.0-88-generic
Operating System: Ubuntu 24.04.3 LTS
OSType: linux
Architecture: x86_64
CPUs: 4
Total Memory: 15.21GiB
Name: athena
ID: c019e0ad-5661-4aa8-a2c5-b518ce5458ef
Docker Root Dir: /var/lib/docker
Debug Mode: false
Experimental: false
Insecure Registries:
::1/128
127.0.0.0/8
Live Restore Enabled: false

Best regards
Gianluca

[chvvkumar]

Hmm what CPU is it using? Can you paste in these logs please?

cat /proc/cpuinfo | grep -E "model name|flags" | head -4

grep -o 'avx[^ ]*' /proc/cpuinfo | head -1

Run the container interactively to see the full error. This will show you the actual Python error/traceback before it crashes.:

docker run -it --rm --network=host \
  -v /home/gianluca/docker_apps/skynet/simpleCloudDetect/keras_model.h5:/app/keras_model.h5 \
  -v /home/gianluca/docker_apps/skynet/simpleCloudDetect/labels.txt:/app/labels.txt \
  -e IMAGE_URL="https://192.168.1.173/indi-allsky/images/latest.jpg" \
  -e MQTT_BROKER="192.168.1.30" \
  -e MQTT_PORT="1883" \
  -e MQTT_DISCOVERY_MODE="legacy" \
  -e MQTT_TOPIC="astro/SimpleCloudDetect" \
  -e DETECT_INTERVAL="60" \
  chvvkumar/simpleclouddetect:latest

Or check with Python directly

docker run -it --rm chvvkumar/simpleclouddetect:latest python3 -c "import tensorflow; print(tensorflow.__version__)"

[giangitesta]

Hi,

he results are as follows:

gianluca@athena:~$ cat /proc/cpuinfo | grep -E "model name|flags" | head -4
model name      : Intel(R) Celeron(R) J4125 CPU @ 2.00GHz
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rdtscp lm constant_tsc art arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc cpuid aperfmperf tsc_known_freq pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3 sdbg cx16 xtpr pdcm sse4_1 sse4_2 x2apic movbe popcnt tsc_deadline_timer aes xsave rdrand lahf_lm 3dnowprefetch cpuid_fault cat_l2 cdp_l2 ssbd ibrs ibpb stibp ibrs_enhanced tpr_shadow flexpriority ept vpid ept_ad fsgsbase tsc_adjust sgx smep erms mpx rdt_a rdseed smap clflushopt intel_pt sha_ni xsaveopt xsavec xgetbv1 xsaves dtherm ida arat pln pts vnmi umip rdpid sgx_lc md_clear arch_capabilities
vmx flags       : vnmi preemption_timer posted_intr invvpid ept_x_only ept_ad ept_1gb flexpriority apicv tsc_offset vtpr mtf vapic ept vpid unrestricted_guest vapic_reg vid ple shadow_vmcs ept_mode_based_exec tsc_scaling
model name      : Intel(R) Celeron(R) J4125 CPU @ 2.00GHz

This instruction does not return output:

gianluca@athena:~$ grep -o 'avx[^ ]*' /proc/cpuinfo | head -1
gianluca@athena:~$

gianluca@athena:~$ docker run -it --rm --network=host \
  -v /home/gianluca/docker_apps/skynet/simpleCloudDetect/keras_model.h5:/app/keras_model.h5 \
  -v /home/gianluca/docker_apps/skynet/simpleCloudDetect/labels.txt:/app/labels.txt \
  -e IMAGE_URL="https://192.168.1.173/indi-allsky/images/latest.jpg" \
  -e MQTT_BROKER="192.168.1.30" \
  -e MQTT_PORT="1883" \
  -e MQTT_DISCOVERY_MODE="legacy" \
  -e MQTT_TOPIC="astro/SimpleCloudDetect" \
  -e DETECT_INTERVAL="60" \
  chvvkumar/simpleclouddetect:latest
Starting Unified Cloud Detection Service...
Converting model if needed...
2025-12-28 17:24:37,524 - INFO - No 'groups' attribute found in model config
Starting Alpaca SafetyMonitor (Waitress + MQTT unified service)...
gianluca@athena:~$

This instruction does not return output:

gianluca@athena:~$ docker run -it --rm chvvkumar/simpleclouddetect:latest python3 -c "import tensorflow; print(tensorflow.__version__)"
gianluca@athena:~$ 

[chvvkumar]

Yup, your CPU does not support AVX instructions and TensorFlow 2.18 requires AVX. Can you try it on a different docker host? Or even on a Pi 4/5?

[giangitesta]

Tomorrow I'll set up a virtual machine and try again. I'll keep you posted.
Do you have any advice for me on how to configure the VM?

Gianluca

[chvvkumar]

Nothing special is required, just make sure the hypervisor CPU supports AVX and the VM hardware has it enabled. I am running it on a docker in a VM running on an i5-8259U without having to mess with any VM settings (i.e. all defaults) under ESXi.

[giangitesta]

Hi,
the container starts up in the virtual machine. There is just one small problem when I try to pass a jpg image file. Basically, the service cannot seem to find the file even though the volume is mounted correctly in /tmp inside the container.

This is the code in compose:

services:
  simpleclouddetect:
    container_name: simple-cloud-detect
    image: chvvkumar/simpleclouddetect:latest
    network_mode: host
    restart: unless-stopped
    environment:
      - IMAGE_URL=file:///tmp/image.jpg
      - MQTT_BROKER=192.168.1.30
      - MQTT_DISCOVERY_MODE=homeassistant
      - DEVICE_ID=clouddetect_001
      - DETECT_INTERVAL=60
    volumes:
      - /home/gianluca/docker_apps/simpleCloudDetect/image.jpg:/tmp/image.jpg
      - /home/gianluca/docker_apps/simpleCloudDetect/keras_model.h5:/app/keras_model.h5
      - /home/gianluca/docker_apps/simpleCloudDetect/labels.txt:/app/labels.txt

This is the app log:

simple-cloud-detect  | 2025-12-30 09:28:34,631 - WARNING - Attempt 1/3 failed: Image file not found: tmp/image.jpg
simple-cloud-detect  | 2025-12-30 09:28:35,846 - WARNING - Attempt 2/3 failed: Image file not found: tmp/image.jpg
simple-cloud-detect  | 2025-12-30 09:28:37,008 - ERROR - Failed to load image from file:///tmp/image.jpg after 3 attempts: Image file not found: tmp/image.jpg
simple-cloud-detect  | 2025-12-30 09:28:37,008 - ERROR - Detection failed: Image file not found: tmp/image.jpg
simple-cloud-detect  | 2025-12-30 09:28:37,008 - ERROR - Error in detection loop: Image file not found: tmp/image.jpg

This is the shell I opened to verify that the file is mounted in /tmp:

gianluca@gianluca-VirtualBox:~/docker_apps/simpleCloudDetect$ docker exec -it simple-cloud-detect /bin/sh
$ ls -la
total 2592
drwxr-xr-x 1 appuser appuser    4096 Dec 30 09:23 .
drwxr-xr-x 1 root    root       4096 Dec 30 09:23 ..
drwxr-xr-x 2 appuser appuser    4096 Dec 30 09:23 __pycache__
-rw-r--r-- 1 appuser appuser     667 Dec 30 09:23 alpaca_config.json
-rw-r--r-- 1 appuser appuser  150260 Dec 20 15:41 alpaca_safety_monitor.py
-rwxr-xr-x 1 appuser appuser    1922 Dec 20 15:41 convert.py
-rwxr-xr-x 1 appuser appuser   14895 Dec 20 15:41 detect.py
-rw-rw-rw- 1 appuser appuser 2457008 Dec 30 09:23 keras_model.h5
-rw-rw-rw- 1 appuser appuser      67 Dec 30 08:25 labels.txt
-rwxr-xr-x 1 appuser appuser     440 Dec 20 15:51 start_services.sh
$ cd /tmp
$ ls -la
total 428
drwxrwxrwt 1 root    root      4096 Dec 30 09:23 .
drwxr-xr-x 1 root    root      4096 Dec 30 09:23 ..
-rw-rw-r-- 1 appuser appuser 426421 Dec 30 09:22 image.jpg

The permissions seem correct to me.
From the log, it seems to ignore the "/" in front of "tmp"; it should be "/tmp/image.jpg" instead of "tmp/image.jpg", but that's just a guess.

Best regards
G

[chvvkumar]

Hi!

Thank you for identifying a bug with the program! The old code used parsed.path.lstrip('/'). In the container, this incorrectly turned the absolute path /tmp/image.jpg into the relative path tmp/image.jpg, which the application couldn't find.

I pushed a change to fix this by removing the unconditional lstrip('/'). On Linux, the code now takes the path /tmp/image.jpg as is, correctly treating it as an absolute path. The if os.name == 'nt' (i.e. if running under Windows), the block is not executed on Linux.

ceccb89

Please pull the image from snd tag and see if it works now.

services:
  simpleclouddetect:
    container_name: simple-cloud-detect
    image: chvvkumar/simpleclouddetect:snd
    network_mode: host
    restart: unless-stopped
    environment:
      - IMAGE_URL=file:///tmp/image.jpg
      - MQTT_BROKER=192.168.1.30
      - MQTT_DISCOVERY_MODE=homeassistant
      - DEVICE_ID=clouddetect_001
      - DETECT_INTERVAL=60
    volumes:
      - /home/gianluca/docker_apps/simpleCloudDetect/image.jpg:/tmp/image.jpg
      - /home/gianluca/docker_apps/simpleCloudDetect/keras_model.h5:/app/keras_model.h5
      - /home/gianluca/docker_apps/simpleCloudDetect/labels.txt:/app/labels.txt

[giangitesta]

Hi,
Thank you for choosing me as a beta tester! I am honoured.

The path now works, but it seems to me that the fix has generated a warning that was not there before (or maybe I didn't see it before):
imple-cloud-detect  | To enable the following instructions: AVX2 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags.
simple-cloud-detect  | 2025-12-30 16:01:50,069 - INFO - Configuration saved to alpaca_config.json
simple-cloud-detect  | /app/alpaca_safety_monitor.py:310: DeprecationWarning: Callback API version 1 is deprecated, update to latest version
simple-cloud-detect  |   client = mqtt.Client()
simple-cloud-detect  | 2025-12-30 16:01:50,214 - INFO - Connected to MQTT broker at 192.168.1.30:1883
simple-cloud-detect  | 2025-12-30 16:01:50,219 - INFO - Pre-loading ML model...
simple-cloud-detect  | /usr/local/lib/python3.11/site-packages/keras/src/layers/core/dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.
simple-cloud-detect  |   super().__init__(activity_regularizer=activity_regularizer, **kwargs)
simple-cloud-detect  | 2025-12-30 16:01:50,636 - INFO - Published HA discovery configurations
simple-cloud-detect  | 2025-12-30 16:01:50,636 - INFO - ML model loaded successfully
simple-cloud-detect  | 2025-12-30 16:01:50,637 - INFO - Published HA discovery configurations
simple-cloud-detect  | 2025-12-30 16:01:50,637 - INFO - Performing initial detection (blocking)...
simple-cloud-detect  | 2025-12-30 16:01:51,463 - INFO - Detection: {'class_name': 'Mostly Cloudy', 'confidence_score': 98.14, 'Detection Time (Seconds)': 0.8, 'image': <PIL.Image.Image image mode=RGB size=1920x1080 at 0x74213E31F290>}
simple-cloud-detect  | 2025-12-30 16:01:51,635 - INFO - Initial detection complete: Mostly Cloudy

However, both MQTT and Alpaca services are up and running. It's amazing!

If I may, one last consideration. I have Indi Allsky running on HTTPS by default (without a valid certificate by default). If you try to retrieve the latest image via the HTTP connector, it errors because it cannot validate the certificate. Perhaps you could set an environment variable to disable the check. I got around this by installing an nginx proxy with a duckdns certificate, but I'm a crazy masochist.

This is the certificate error:

simple-cloud-detect  | 2025-12-30 15:55:00,686 - WARNING - Attempt 1/3 failed: HTTPSConnectionPool(host='192.168.1.173', port=443): Max retries exceeded with url: /indi-allsky/images/latest.jpg (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: self-signed certificate (_ssl.c:1016)')))
simple-cloud-detect  | 2025-12-30 15:55:01,732 - WARNING - Attempt 2/3 failed: HTTPSConnectionPool(host='192.168.1.173', port=443): Max retries exceeded with url: /indi-allsky/images/latest.jpg (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: self-signed certificate (_ssl.c:1016)')))
simple-cloud-detect  | 2025-12-30 15:55:02,916 - ERROR - Failed to load image from https://192.168.1.173/indi-allsky/images/latest.jpg after 3 attempts: HTTPSConnectionPool(host='192.168.1.173', port=443): Max retries exceeded with url: /indi-allsky/images/latest.jpg (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: self-signed certificate (_ssl.c:1016)')))
simple-cloud-detect  | 2025-12-30 15:55:02,916 - ERROR - Detection failed: HTTPSConnectionPool(host='192.168.1.173', port=443): Max retries exceeded with url: /indi-allsky/images/latest.jpg (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: self-signed certificate (_ssl.c:1016)')))
simple-cloud-detect  | 2025-12-30 15:55:02,957 - ERROR - Initial detection failed: HTTPSConnectionPool(host='192.168.1.173', port=443): Max retries exceeded with url: /indi-allsky/images/latest.jpg (Caused by SSLError(SSLCertVerificationError(1, '[SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: self-signed certificate (_ssl.c:1016)')))

Best regards
G

[chvvkumar]

Haha, thank you for your patience while getting this working. I use a URL so almost never used the file option apart from initial tests.

As for the cert check, let me see if I can add an ignore cert option per your suggestion.

[chvvkumar]

And oh, I have mine behind a reverse proxy with letsencrypt certs as well so definitely feel you on the masochistic part.