> ## Documentation Index
> Fetch the complete documentation index at: https://docs.salad.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Job Queue Autoscaling Overview

> Learn how SaladCloud Job Queue autoscaling automatically adjusts container instances based on queue depth to optimize performance and cost for batch processing workloads.

*Last Updated: July 21, 2025*

**Job Queue Autoscaling** in SaladCloud dynamically adjusts the number of container instances based on the number of
jobs in your SaladCloud job queues. This ensures that your infrastructure scales up when job volume increases and scales
down during low usage periods, optimizing both performance and cost efficiency.

Job Queue autoscaling integrates directly with SaladCloud's managed queue service and the Job Queue Worker to provide
seamless scaling for batch processing and job-based workloads.

## Prerequisites

Before you can enable Job Queue autoscaling, you must configure the Job Queue:

1. Ensure that [**SaladCloud Job Queues**](/container-engine/explanation/job-processing/job-queues) are enabled for your
   container group.
2. Verify that your application can handle scaling with multiple container instances.
3. Review the [Job Queue Autoscaling Settings](/container-engine/reference/autoscaling/settings) to understand how to
   fine-tune autoscaling for your workloads.

## Example Use Case

Let's say you run a batch processing workload, but the number of jobs you receive is inconsistent. Sometimes, there are
no jobs at all, and you want to scale down to zero instances to avoid paying for unused compute resources. By setting
the **Minimum Replicas** to 0, the Job Queue autoscaler will automatically scale down to zero when the queue is empty,
which means you don't pay for any resources when there is no workload. However, this will result in cold starts when new
jobs arrive. Cold starts occur when the system needs to initialize new instances, which can introduce a delay while
containers are started from an inactive state.

At other times, you might experience a spike where thousands of jobs need to be processed quickly. By setting the
**Maximum Replicas** to 100, you can scale up and process all the jobs much faster. Even though you're using more
instances, you're still paying for the same amount of compute hours you would have spent on one instance, but the jobs
are completed 100 times faster, reducing overall processing time.

***

Ready to configure Job Queue autoscaling? Learn more about the available
[Settings](/container-engine/reference/autoscaling/settings) or
[Set Up Job Queue Autoscaling](/container-engine/how-to-guides/autoscaling/enable-autoscaling) in SaladCloud.