Render is a SaaS which makes it really easy to deploy custom Docker-based services around the world with a few clicks. Faktory is a persistent background job system for any programming language. Let's get a Faktory instance running for your app.
- Sign in at Render.com.
- On your Dashboard, select "New Private Service".
- Paste in the URL for this repo:
https://github.com/contribsys/faktory-render and click the box that
appears.
- Customize it:
- Enter a unique name for your Faktory server: "faktory", "foobar-faktory", etc.
- Select an appropriate region "near" your app.
- Notice you can set ENV variables or secrets files, useful for FAKTORY_LICENSE if you upgrade to Faktory Pro or Enterprise.
- Change the value for FAKTORY_PASSWORD in Environment Variables.
- Select an appropriate instance size and create a persistent disk for your data, mounted at /var/lib/faktory. See sizing notes below.
- Click "Create Private Service". You should have a live Faktory instance in a minute. Both ports should be accessible to your app running in Render.
Faktory doesn't need much disk because queues are typically empty. Only your error queue and scheduled queue will contain long-term data so I recommend going small: 5-10GB should be plenty of space for most people.
Faktory uses Redis under the covers and all data must fit into RAM with Redis. A small, test instance can run with 1GB. A light instance processing less than 10m jobs/day can run with 2GB. For bigger users or users using more data-heavy features like Batches, go with 4GB. Likewise, go with 1, 2 or 4 CPUs based on your expected usage.
Remember we are spec'ing out the instance running Faktory only. It's your worker instances that will actually execute jobs: that's where you'll spend most of your resource budget.
Unfortunately private services are just that: private. You can't use a Faktory service deployed to render.com with a generic app deployed in AWS because there is no access to the service via public network. If this changes, I will update this section with details.