# Run concepts

## Deployment

Deployment refers to the process of activating pipelines that have already been configured with a trigger. This activation can happen in either test or production. If you want to learn more about how to deploy a pipeline, read the [documentation](https://docs.digibee.com/documentation/developer-guide/development-cycle/overview/deployment/deployments).<br>

The deployment has three related concepts:&#x20;

### Pipeline size

The pipeline deployment size depends on the processing capacity and memory of each replica. There are three sizes for deployments:

<figure><img src="https://content.gitbook.com/content/jvO5S91EQURCEhbZOuuZ/blobs/r23qSIVS33WDeWqk789g/RunConcepts_PipelineSize.png" alt=""><figcaption></figcaption></figure>

* SMALL, 10 consumers, 20% of 1 CPU, 64 MB memory&#x20;
* MEDIUM, 20 consumers, 40% of 1 CPU, 128 MB memory&#x20;
* LARGE, 40 consumers, 80% of 1 CPU, 256 MB memory

For example, if you configure a SMALL deployment size for your pipeline executions, it means that 10 messages can be concurrently processed.

### Concurrent executions

Concurrent executions represent the number of executions each deployed replica can run at the same time. The maximum number of concurrent executions depends on the deployment size configured.

<figure><img src="https://content.gitbook.com/content/jvO5S91EQURCEhbZOuuZ/blobs/nYQ60v9YaLTvXCo1XWVR/RunConcepts_ConcurrentExecutions.png" alt=""><figcaption></figcaption></figure>

Learn more about these concepts in the [Pipeline Engine documentation](https://docs.digibee.com/documentation/developer-guide/development-cycle/overview/runtime/pipeline-engine).&#x20;

### Replicas

Choosing the adequate number of replicas will support your integrations with high availability, guaranteeing their autonomy and the efficiency of their concurrent executions.&#x20;

<figure><img src="https://content.gitbook.com/content/jvO5S91EQURCEhbZOuuZ/blobs/7FwzcVEXf4NSgEv2I3tu/RunConcepts_Replicas.png" alt=""><figcaption></figcaption></figure>