This is fixed with latest package release now.

Since the last update (v1.85.1), the machine learning do not work: [Nest] 42 - 04/02/2025, 5:28:49 PM ERROR [Microservices:{"id":"73e5da08-c8e7-4d1a-b557-e210f2cb5289"}] Unable to run job handler (faceDetection/face-detection): Error: Machine learning request '{"facial-recognition":{"detection":{"modelName":"buffalo_l","options":{"minScore":0.7}},"recognition":{"modelName":"buffalo_l"}}}' failed for all URLs Apr 02 19:28:49 Error: Machine learning request '{"facial-recognition":{"detection":{"modelName":"buffalo_l","options":{"minScore":0.7}},"recognition":{"modelName":"buffalo_l"}}}' failed for all URLs Apr 02 19:28:49 at MachineLearningRepository.predict (/usr/src/app/dist/repositories/machine-learning.repository.js:98:15) Apr 02 19:28:49 at process.processTicksAndRejections (node:internal/process/task_queues:105:5) Apr 02 19:28:49 at async MachineLearningRepository.detectFaces (/usr/src/app/dist/repositories/machine-learning.repository.js:107:26) Apr 02 19:28:49 at async PersonService.handleDetectFaces (/usr/src/app/dist/services/person.service.js:239:52) Apr 02 19:28:49 at async JobService.onJobStart (/usr/src/app/dist/services/job.service.js:156:28) Apr 02 19:28:49 at async EventRepository.onEvent (/usr/src/app/dist/repositories/event.repository.js:126:13) Apr 02 19:28:49 at async Worker.processJob (/usr/src/app/node_modules/bullmq/dist/cjs/classes/worker.js:394:28) Apr 02 19:28:49 at async Worker.retryIfFailed (/usr/src/app/node_modules/bullmq/dist/cjs/classes/worker.js:581:24) Apr 02 19:28:49 [Nest] 42 - 04/02/2025, 5:28:49 PM WARN [Microservices:MachineLearningRepository] Machine learning request to "http://127.0.0.1:3003" failed: fetch failed If I try to start it from supervisord: Apr 02 19:36:12 2025-04-02 17:36:12,166 INFO spawned: 'machine-learning' with pid 520 Apr 02 19:36:12 warning: `VIRTUAL_ENV=/app/code/venv` does not match the project environment path `.venv` and will be ignored; use `--active` to target the active environment instead Apr 02 19:36:12 error: Failed to install: httpx-0.28.1-py3-none-any.whl (httpx==0.28.1) Apr 02 19:36:12 Caused by: failed to create directory `/app/code/machine-learning/.venv/lib/python3.12/site-packages/httpx`: Read-only file system (os error 30) Apr 02 19:36:12 2025-04-02 17:36:12,247 WARN exited: machine-learning (exit status 2; not expected) Also, I'm not sure if this occurs since the last upgrade: root@2e57dda8-6002-47bd-b15c-be4345bc5a94:/usr/src/app# admin-cli node:internal/modules/cjs/loader:1228 throw err; ^ Error: Cannot find module '/app/code/server/dist/main' at Function._resolveFilename (node:internal/modules/cjs/loader:1225:15) at Function._load (node:internal/modules/cjs/loader:1055:27) at TracingChannel.traceSync (node:diagnostics_channel:322:14) at wrapModuleLoad (node:internal/modules/cjs/loader:220:24) at Function.executeUserEntryPoint [as runMain] (node:internal/modules/run_main:170:5) at node:internal/main/run_main_module:36:49 { code: 'MODULE_NOT_FOUND', requireStack: [] } Node.js v22.14.0

I'm already using a volume for the main data storage of Nextcloud. My idea was to mount this volume in the Immich app and let it index the pictures inside using an external library. I was not aware of the existence of the "media" group. Once I applied it to the data directory of Nextcloud (chown :media data/), Immich was able to crawl the directory! Thanks!

Also see: https://docs.cloudron.io/apps/immich/#settings After changing the file, the app needs to be restarted.

I would not recommend implementing a custom backup scheme, easy to miss things. If you use some sshfs rsync backup storage with hardlinks, the backups would be quite efficient. For extra safety the snapshotting of course helps, but anything custom will cause more trouble if things go wrong in the future.

@nebulon said in HEIC Files: A new package with a patched libheif is out now. Here at least the sample photos we got work as expected now. I'll check, fingers crossed! Thanks for this! Further to this, I just tested - It's fixed! Perfect. Thanks for the quick resolution.

@mdc773 you mean https://www.idrive.com/s3-storage-e2/ ? I will give it a try, thanks for the suggestion. @nebulon yes. I am using a vpn, maybe it's linked to that.

The tests are failing - https://git.cloudron.io/packages/immich-app/-/merge_requests/26 . We are investigating

Glad it worked out. Might be worth to report upstream in the Immich issue tracker actually.

Hi, thanks for answer. As my VPS was "young", I erased everything et re-installed a clean ubuntu on it, then I replaced Cloudflare by bunny. Looks like it works good now.

Hi, thanks for answer. It's strange because despite the popup, updates are ok (location, times, etc).

This is OK now. I think that at some point I rebooted that computer and the issue was fixed. I will keep your suggestion in mind for next time. Thanks!

@girish Thank you, I'll try writing there and let you know.

I found an solution to this issue by switching to rsync for backups (with hetzner storage box). Since each backup is incremental, it doesn't matter how many backups are stored. The backup size pretty much matches the app size.

Circling back after some testing. Rsync with backblaze didn't seem to perform very well. Tho i'm not sure i had it configured correctly. It looked like it tried to sync everything each time. Rsync set up with hetzner storage box works like a charm. Each sync after the first one is fast.

This fixed, thank you, @kubasbimbas, I appreciate it.

Indeed, I've pushed a package patch version, please check for updates on immich again and this time it contains the correct 1.117.0

I am not quite sure where those errors are from and why its a screenshot. However: can you check if the redis container together with immich is up and running? seems to be an issue due to 1) This is an error I can reproduce and will investigate, however it has nothing to do with the admin creation issue you seem to have Looking at the folder name, this comes from the host system and not the immich app instance? Is this while uninstalling or updating the app?