--- title: 'Horizon - Beautiful queue monitoring' source: 'https://youtube.com/watch?v=Lb6V4yQ2B4s' video_id: 'Lb6V4yQ2B4s' date: 2026-06-14 duration_sec: 304 --- # Horizon - Beautiful queue monitoring > Source: [Horizon - Beautiful queue monitoring](https://youtube.com/watch?v=Lb6V4yQ2B4s) ## Summary Laravel Horizon provides a beautiful dashboard and code-driven configuration for managing Redis queues. It allows monitoring key metrics like job throughput, runtime, and failures, and enables easy scaling and configuration of workers. ### Key Points - **Introduction to Horizon** [0:00] — Horizon is a powerful tool for managing Redis queues with a dashboard and code-driven configuration. - **Monitoring Metrics** [0:16] — Horizon monitors job throughput, runtime, job failures, and other key metrics. - **Central Hub** [0:28] — Horizon acts as a central hub for all queue monitoring needs, with a single configuration file for workers. - **Dashboard Access** [0:41] — Access the dashboard via /horizon; it shows active status, supervisors, and job data. - **Starting Horizon** [1:00] — Run 'php artisan horizon' to start Horizon and push jobs to the queue. - **Tweaking Config** [1:58] — Increase max processes in the config file to speed up job processing. - **Auto-balancing** [2:29] — Horizon auto-balances processes across queues, allocating more to queues with more jobs. - **Finding Right Settings** [3:14] — Adjust processes, retries, and timeouts based on your specific needs. - **Summary** [4:39] — Horizon is essential for effective queue management, providing insights and enhancing background job processing. ### Conclusion Laravel Horizon is essential for managing queues effectively, offering a dashboard and auto-balancing to optimize job processing. Check the official docs for advanced use. ## Transcript Are you ready to take your Q management to the next level? Let's explore Level Horizon, a powerful tool that provides a beautiful dashboard and code-driven configuration for managing your Rady's cues. With Horizon, you can easily monitor key metrics of your Q system, such as job throughput, runtime, job failures, and so on. This means you can keep your application running smoothly and efficiently while gaining valuable insights into your background jobs. Horizon acts as a central hub for all your Q monitoring [music] needs. It allows you to configure your workers in a single manageable configuration file, making it easy to scale and modify your co-workers as needed. Let's take a look together. Once you have Horizon installed, you can access the dashboard by going to / Horizon application. And this is what it looks by looks like by default. We don't have any data yet. And we can see that it's currently inactive. So let's change this. Now let's start Horizon and push some jobs to the queue. And we can do this by running PHP artisan horizon. Here we go. And you can also see I have my Q worker for the default queue already running. So when we go back to the dashboard, we can now see it's active. We also see some more information here about the queue about supervisors. And we're going to talk about this just in a bit. But let's push some chops here so that we actually see something. And in our application which is called user lift, we're going to import some users. And then we're going to take a look again. And you can see suddenly something is happening here. We have 100 and one chops. Here we see jobs per minute. We see currently it's just running one process. And we can see here on the left the pending jobs. It seems like everything was already running. And here are our completed jobs which we used in order to import all the users to our import users table. This already went pretty well, but we can speed things up a little bit while tweaking our Horizon config file. Now, let's increase the number of max processes here for our local supervisor here, which we're using. Then, we're going to need to rerun Horizon and rerun our Qwork. And then back here now, we're going to import, let's say, 1,000 users here. And if we go to the dashboard, let's see what's happening now here. Okay, we can already see that there are one around 1,000 chops here being added to our default queue. And we can see it's already using all the three processes. And you can also see that chops per minute is now way faster than it was before. So this means more processes of course means that you can handle more chops at the same time. And I think we're already done here. Let's see. Now we still have some pending jobs here. The completed ones. You could also see failed ones, but we don't have any of them here. But it looks like it's still running here. But it's going pretty fast here as you can see. And I think we should be already done in a couple of seconds. 80 more jobs here. Here we go. Pending jobs. No, we're done now. And yeah, this was of course now very fast now while running while using three processes. And you can see it has has gone down again to one process again because that's the only thing we currently need because there are no chops on our default queue. It is important to mention that you have to find the right settings, amount of processing, retries or timeouts for your own situation. There is a lot you can tweak to your exact needs. But I want to show you here one more thing. Okay, back here in the config file for our local supervisor, we have defined the default queue, a male ske, and the backups Q. And we have given them 10 processes to share. And now again, let's import let's say 1,000 users again. Let's start this and let's go back here. We can already see we have 1,000 chops here, but we also see the other cues here, backups on meals. We have no chops here at all. And you can see it's already using a little more processes than before, now it's using um five processes. And our supervisor is currently using already eight. So this means from those 10 processes which we have allocated for the supervisor one now most of them are being used by default Q because they already see that there are no shops being processed on backups and mails and this is called autobalancing. So that it use all the possible processes for the [snorts] queue which has the most chops and you can see it's already done by processing around 1,000 chops per minute. And now the processes went back to just one because yeah, Horizon sees there that there are zero jobs for all the other cues. So only three processes are currently being used for our supervisor. In summary, Laval Horizon is essential tool for any Laval developer looking to manage their cues effectively. It enhances your application background job processing capabilities and provides valuable insights into your job performance. For more details and advanced uses, be sure to check out the official docs. As always, happy coding and enjoy managing your cues with Larable Horizon.