ARQ¶

ARQ is for async Redis worker deployments. Wove enqueues a named ARQ function with the payload, and that function posts completion back through Wove’s callback URL.

Use When¶

Your project already runs ARQ workers.
Worker code is async and Redis-backed.
You want a named worker function to execute Wove payloads.

Execution Shape¶

Wove creates or uses an ARQ pool.
Wove calls enqueue_job(function_name, payload, ...).
The ARQ worker function calls await wove.integrations.worker.arun(payload).
The worker posts Wove completion events back to the callback URL.

Dependency¶

Install dispatch support and ARQ in the submitting process and in ARQ workers.

pip install "wove[dispatch]" arq

Configure Wove¶

import wove
from arq.connections import RedisSettings

wove.config(
    default_environment="arq",
    environments={
        "arq": {
            "executor": "arq",
            "executor_config": {
                "redis_settings": RedisSettings(host="redis"),
                "function_name": "myapp_wove_task",
                "queue_name": "wove",
                "callback_token": "shared-secret",
                "callback_url": "https://wove-runner.internal/wove/events/shared-secret",
            },
        }
    },
)

You can pass an existing ARQ pool as executor_config["pool"]. If pool is omitted, Wove creates one with redis_settings.

Worker Function¶

from wove.integrations.worker import arun


async def myapp_wove_task(ctx, payload):
    return await arun(payload)


class WorkerSettings:
    functions = [myapp_wove_task]
    queue_name = "wove"

Options¶

Key	Effect
`pool`	Existing ARQ Redis pool.
`redis_settings`	Settings used when Wove creates the pool.
`function_name`	ARQ function name. Defaults to `wove_run_backend_payload`.
`queue_name`	Optional ARQ queue name.
`enqueue_options`	Extra keyword arguments passed to `enqueue_job(...)`.