How to improve AI task pickup speed in AIO

[szaimen]

Most AI tasks will be run as part of the background job system in Nextcloud which only runs jobs every 5 minutes by default. To pick up scheduled jobs faster you can set up background job workers that process AI tasks as soon as they are scheduled. If the PHP code or the Nextcloud settings values are changed while a worker is running, those changes won’t be effective inside the runner. For that reason, the worker needs to be restarted regularly. It is done with a timeout of N seconds which means any changes to the settings or the code will be picked up after N seconds (worst case scenario). This timeout does not, in any way, affect the processing or the timeout of the AI tasks.

Use these instructions https://docs.nextcloud.com/server/latest/admin_manual/ai/overview.html#systemd-service but in step 2, use the following script:

#!/bin/bash
echo "Starting Nextcloud AI Worker $1"

# Function to attempt starting the worker
start_worker() {
    echo "Attempting to start Nextcloud AI Worker $1..."
    
    # Run the worker inside the container (60 seconds for task execution)
    docker exec --user www-data nextcloud-aio-nextcloud php occ background-job:worker -t 60 'OC\TaskProcessing\SynchronousBackgroundJob'
    
    # Capture the exit status of the docker exec command
    exit_code=$?

    # Check if the worker ran successfully
    if [ $exit_code -eq 0 ]; then
        echo "Worker completed successfully."
        return 0  # Success
    else
        echo "Worker failed to start or complete with exit code $exit_code."
        return 1  # Failure
    fi
}

# Loop to start the worker and retry if it fails
while true; do
    # Check if the container is running
    if ! docker ps --format "{{.Names}}" | grep -q "^nextcloud-aio-nextcloud$"; then
        echo "Nextcloud container is not running. Waiting 60 seconds before next attempt."
        sleep 60  # If the container isn't running, wait for 60 seconds
        continue  # Skip to the next iteration
    fi

    # Start the worker
    start_worker

    # Check if the worker succeeded or failed
    if [ $? -eq 1 ]; then
        # If failed, wait 30 seconds (or adjust as needed) to avoid rapid retries
        echo "Worker failed. Waiting 30 seconds before retrying..."
        sleep 30
    fi
done

[purelyillusive]

what container to run this in?

[szaimen]

Please read the instructions, especially the linked docs.

[derStephan]

I read the instructions but don't really understand what to do for AIO.

After running

occ background-job:worker -v -t 60 "OC\TaskProcessing\SynchronousBackgroundJob"

it really speeds up AI-things until the next reboot of the host server.

The I tried adding

@reboot set -e; while true; do docker exec -it nextcloud-aio-nextcloud sudo -E -u www-data php occ background-job:worker -v -t 60 "OC\TaskProcessing\SynchronousBackgroundJob"; done

But that did not work. Can you explain how to set that option persistently? I can send a PR for the official documentation afterwards.

[szaimen]

You need to create a systemd service as mentioned in the docs...

[derStephan]

Ok, so I was mislead by the sentence

For Nextcloud-AIO you should use this command on the host server.

[derStephan]

Ok, here is what I have done for AIO on the physical host:

sudo bash
nano /root/taskprocessing.sh

put the content of @szaimen in and save the file, then do

chmod +x /root/taskprocessing.sh
crontab -e

append the following and save

@reboot /root/taskprocessing.sh

This works for me on Ubuntu 24.04 without adding systemd service

[purelyillusive]

that's actually a good solution though. Because nextcloud aio does not have systemd(at least when i tried to use it) that could actually work, unlike the systemd way of doing it Sent with [Proton Mail](https://proton.me/mail/home) secure email.

…

On Wednesday, July 2nd, 2025 at 1:40 AM, derStephan ***@***.***> wrote: Ok, here is what I have done for AIO on the physical host: sudo bash nano /root/taskprocessing.sh put the content of ***@***.***(https://github.com/szaimen) in and save the file, then do chmod +x /root/taskprocessing.sh crontab -e append the following and save @reboot /root/taskprocessing.sh This works for me on Ubuntu 24.04 without adding systemd service — Reply to this email directly, [view it on GitHub](#5430 (reply in thread)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/BRQHGN2H6JVEVKTM7F53TKT3GOLGZAVCNFSM6AAAAACAHZ5XBGVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGNRTHAYTMMQ). You are receiving this because you commented.Message ID: ***@***.***>

[derStephan]

AIO does not have to have systemd. This is to be done on the host.

[purelyillusive]

ohhhhhh. well that makes much more sense now. Sent with [Proton Mail](https://proton.me/mail/home) secure email.

…

On Thursday, July 3rd, 2025 at 1:04 AM, derStephan ***@***.***> wrote: AIO does not have to have systemd. This is to be done on the host. — Reply to this email directly, [view it on GitHub](#5430 (reply in thread)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/BRQHGN2TEBUIIXYD3BIX3V33GTPYDAVCNFSM6AAAAACAHZ5XBGVHI2DSMVQWIX3LMV43URDJONRXK43TNFXW4Q3PNVWWK3TUHMYTGNRUHEZDONQ). You are receiving this because you commented.Message ID: ***@***.***>