No images being sent, no increased hardware usage seen

[CombFiltered]

I followed all the instructions and enabled the bot, and added it to the Talk conversation. I copied the command and even tried manually typing @image just in case it was a specific tag (it doesn't seem to do anything special), and no image pops up. I do not see any increase in memory or CPU usage on the VM.

I do not have a discrete GPU installed with the VM running Nextcloud, and I did not see a way to configure it for CPU vs GPU. I do see in the daemon that the GPU support is false, so I'm assuming it interpreted no discrete GPU properly during setup?

I do not see it throwing anything in the general Nextcloud logs. Is there another place I can check to see what's happening (or not happening)?

The Nextcloud VM has plenty of CPU overhead, and has been assigned 28GB of memory. Just under 8GB is in use currently, so per the documentation that should be plenty of RAM to run the CPU modeling. I wanted to see how the CPU modeling ran before I tried to add a discrete GPU pass thru.

[bigcat88]

1. Do you have 2FA enabled?
2. docker logs XYZ > container.log 2>&1

where XYZ name of container with app and please attach these logs.

[andrey18106]

so I'm assuming it interpreted no discrete GPU properly during setup?

Default daemon is configured without GPUs support. If your daemon has NVIDIA GPUs - you can create another daemon with checkbox "Enable GPUs support" in deploy config enabled, after that ExApps containers should be created with forwarded GPUs access.

[CombFiltered]

1. Do you have 2FA enabled?
2. docker logs XYZ > container.log 2>&1

where XYZ name of container with app and please attach these logs.

I do not have 2FA enabled.

I looked through all the logs to see which one might obviously contain logs from the app, and saw none. I apologize for the ignorance, but I'm running the standard Nextcloud docker install. Which would it be? I have containers:
nextcloud-aio-apache
nextcloud-aio-notify-push
nextcloud-aio-nextcloud
nextcloud-aio-docker-socket-proxy
nextcloud-aio-imaginary
nextcloud-aio-talk-recording
nextcloud-aio-fulltextsearch
nextcloud-aio-clamav
nextcloud-aio-redis
nextcloud-aio-database
nextcloud-aio-talk
nextcloud-aio-collabora
nextcloud-aio-mastercontainer

[andrey18106]

It should be nc_app_ai_image_generator_bot container name.

[CombFiltered]

Got it, for some reason that container did not come up when I listed them
image-container.log

[bigcat88]

can we see Nextcloud.log as well?

[bigcat88]

sorry, no need for log, it is a strange error, I will try to reproduce tomorrow

please say what CPU model do you have and does it virtualized or not

[CombFiltered]

Great, thanks.

This is running on a server with dual Xeon E5-2690 v4 CPUs, and is virtualized with Hyper-V.

[bigcat88]

I will try to reproduce tomorrow

Sorry for the late answer, I have Nextcloud on Unraid at home, I tried to install AppAPI so that I could write a guide on how to do it right away, but I encountered difficulties as Undraid does not map docker socket inside NC container(I do no want to do that with hands), we will try to solve this in a week or two, and only after that I will be able to check..

[CombFiltered]

No worries. AppAPI is installed by default it seems with the AIO Docker install. At least I didn't install it, so it may have been auto installed with another optional container that I installed. The Docker Socket Proxy had to be installed separate. Let me know if there's anything I can help investigate or try.

[bigcat88]

Unfortunately, on my home Unraid with Intel® Core™ i7-9700K CPU without GPU I can't reproduce an error.

You can give it a try, but we did not change in AppAPI 2.0 anything related to this, but who knows..

[CombFiltered]

It does not appear to be any different. Is there anything I can check or log better for you?

[bigcat88]

AppAPI should be 2.0.0+ version and ImageBot 2.0 version.

Archive and attach container.log file, that you can generate with this command:
docker logs nc_app_ai_image_generator_bot > container.log 2>&1

Output of this command will ensure us that it CPU supports all commands(we are interested in flags):
docker exec nc_app_ai_image_generator_bot lscpu

[CombFiltered]

Hopefully this will lead us somewhere

container.log

[bigcat88]

Oh, the download speed is “52.4 KB/s”... our current initialisation algorithm in AppAPI will not work in this case, since the loading of AI models will not complete in 40 minutes...

@andrey18106 we need ex_app_init_timeoutconfig key to be configurable in AppAPI settings UI