> ## Documentation Index
> Fetch the complete documentation index at: https://docs.automq.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Manage Connectors
> Create a Connect Cluster, create Connectors on it, and manage Connector lifecycle operations.
AutoMQ Connect uses two resource types. A **Connect Cluster** provides the Kafka Connect worker runtime and compute capacity. A **Connector** defines a source or sink data integration workload that runs on a Connect Cluster.
## Prerequisites
Before you begin, prepare the following resources:
* A running AutoMQ instance.
* Kafka credentials with permissions for the topics and consumer groups that the Connector will use. For more information, see [Manage Kafka ACLs](/automq-cloud/manage-security/manage-kafka-acls).
* The source or sink system and its access credentials.
* A Kubernetes cluster, namespace, and ServiceAccount for the Connect Cluster workers. On AWS, bind the ServiceAccount to an IAM role that grants access to the Ops Bucket and any external AWS resources required by the plugin. For EKS preparation instructions, see [Deploy to Amazon EKS](/automq-cloud/deploy-automq-on-kubernetes/deploy-to-aws-eks).
Grant only the permissions required by the workload. A Sink Connector requires read permission for its Kafka topics and consumer group. A Source Connector requires write permission for its destination topics.
## Create a Connect Cluster
1. Sign in to the AutoMQ Console.
2. In the navigation pane, choose **Kafka Connect**, choose the **Connect Clusters** tab, and then choose **Create Connect Cluster**.
3. Configure the Connect Cluster parameters.
| Parameter | Description |
| ----------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------- |
| Name | A name that is unique within the current environment. |
| Description | An optional description of the worker cluster. |
| Version | The Kafka Connect runtime version. |
| Plugins | The Managed or custom plugin versions and connector classes to install on the worker cluster. |
| AutoMQ Instance | The AutoMQ instance used by the Connect Cluster. |
| Number of Workers | The number of Kafka Connect workers. Valid values are 1 through 200. |
| Worker Tier | The CPU and memory allocated to each worker: 0.5 vCPU and 1 GiB, 1 vCPU and 2 GiB, 2 vCPU and 4 GiB, or 4 vCPU and 8 GiB. |
| Kubernetes Cluster | The Kubernetes cluster where the workers run. |
| Namespace | The Kubernetes namespace for the workers. |
| ServiceAccount | The Kubernetes ServiceAccount used by the worker Pods. |
| Cloud Identity | The cloud identity associated with the ServiceAccount, such as an AWS IAM role. It must grant access to the cloud resources required by the plugins. |
| SchedulingSpec | Optional Kubernetes scheduling settings, including node selectors, affinity and anti-affinity rules, and tolerations. |
| Worker Configuration | Optional Kafka Connect worker properties. |
| Prometheus Remote Write | Optional endpoint and authentication settings for pushing Connect Cluster metrics to Prometheus-compatible storage. |
4. Review the settings and choose **Create**. Wait until the Connect Cluster status is **Running** before creating a Connector.
Ensure that the selected Kubernetes node pool has enough allocatable CPU and memory for the worker count and Worker Tier that you select.
## Create a Connector
1. In **Kafka Connect**, choose the **Connectors** tab and then choose **Create Connector**.
2. Configure the Connector parameters.
| Parameter | Description |
| -------------------- | -------------------------------------------------------------------------------------------------- |
| Connect Cluster | The running Connect Cluster on which the Connector runs. |
| Name | A name that is unique within the selected Connect Cluster. |
| Description | An optional description of the data integration workload. |
| Plugin | A plugin installed on the selected Connect Cluster. |
| Connector Type | Source or Sink. The available type depends on the selected plugin. |
| Connector Class | The connector implementation class provided by the plugin. |
| Task Count | The maximum number of Connector tasks that can run in parallel. |
| Kafka Authentication | The Kafka security protocol and credentials used to connect to the AutoMQ instance. |
| Sink Topics | For a Sink Connector, the topics to consume and, when applicable, the consumer group. |
| Source Offsets | For a Source Connector, the offset topic and initial offset settings when supported by the plugin. |
| Plugin Configuration | Connection information and connector-specific properties for the external system. |
3. Enter the plugin configuration. Use the structured form when it is available, or enter key-value properties in custom mode.
4. Review all parameters and choose **Create**.
## Manage a Connector
Open a Connector details page to view its status, tasks, metrics, and logs. You can update supported plugin properties and the task count, and you can perform the following lifecycle operations:
* **Pause:** Stops data processing without deleting the Connector.
* **Resume:** Resumes a paused Connector.
* **Restart:** Restarts the Connector and its tasks to recover from an error or apply supported runtime changes.
Worker count, Worker Tier, Kubernetes placement, installed plugins, and Prometheus Remote Write settings belong to the Connect Cluster. Update those settings on the Connect Cluster rather than on an individual Connector.
## Delete a Connector
On the **Connectors** tab, locate the Connector, choose **Delete**, and confirm the operation.
Deleting a Connector cannot be undone. Verify the Connector name before you confirm the operation.