Install and download agent - Precisely Data Integrity Suite

Data Integrity Suite
Product
Spatial_Analytics
Data_Integration
Data_Enrichment
Data_Governance
Precisely_Data_Integrity_Suite
geo_addressing_1
Data_Observability
Data_Quality
dis_core_foundation
Services
Spatial Analytics
Data Integration
Data Enrichment
Data Governance
Geo Addressing
Data Observability
Data Quality
Core Foundation
ft:title
Data Integrity Suite
ft:locale
en-US
PublicationType
pt_product_guide
copyrightfirst
2000
copyrightlast
2026

To connect your on-premises environment with the Precisely Cloud, begin by downloading the agent installation package from the authorized source. Run the installer and follow the prompts to complete setup. During installation, configure required settings such as access credentials and network parameters. After installation, apply any additional configurations needed. A properly installed and configured agent is essential for secure and reliable communication with the Precisely Cloud.

Install an agent

The Agent enables execution of Data Integration, Data Replication, profiling, and quality workloads within your infrastructure.

The Agent can be deployed:

  • On a single machine in Standalone mode(Default installation mode).

  • On an existing Kubernetes cluster in Cluster mode.

Installation is performed using the Agent CLI (discli), which is generated per agent instance from the platform UI.

System requirements

System requirements Description
Supported Operating Systems

For standalone mode: Ubuntu 24.04 and Red Hat Enterprise Linux 10 (RHEL 10), (Linux Kernel 4.15+)

For cluster mode: Ubuntu 24.04 and Red Hat Enterprise Linux 10 (RHEL 10), Amazon Linux 2023

Note: Ubuntu 22.04 LTS and RHEL 9.x remain supported, along with the new versions listed above.
Supported Hardware Any system with x86_64 architecture
Resources
  • vCPU >= 8
  • RAM >= 16GB
    • For On-prem profiling: 64GB free
  • Hard disk >= 64 GB free
    • Storage for Data Integration or Data Replication: 300GB
Configurations sudo privilege
  • It is recommended to have 10% free space at all times when the agent is running.
  • Agent CLI is built using graalVM. All dependencies such as JRE, and other packages are bundled together.
    Note:

    1. For standalone mode, follow the system requirements listed in Table: System Requirements.

    2. For cluster mode, ensure your Kubernetes cluster meets similar overall capacity as the standalone requirements but distributed across nodes.

    3. You need sudo access to perform all Agent CLI operations, but do not use the root user or prefix any Agent CLI commands with 'sudo', such as 'sudo ./discli version'

Namespace installation behavior

Agent installation behavior has been updated to simplify deployment and management. Agents now install in single-namespace mode by default, rather than multi-namespace mode.

  • Default installation (single-namespace mode)

    When you install an agent without specifying a namespace, it deploys to the default namespace: precisely-ns.

    To specify a custom namespace, use the -n flag: ./discli agent install -n my-namespace

  • Multi-namespace installation

    If you need to install an agent in multi-namespace mode, you must explicitly enable it using one of these flags:

    • -x true
    • --multi true

    Examples:./discli agent install -x true or ./discli agent install --multi true

Note: Multi-namespace installation is no longer the default behavior. You must explicitly enable it using either -x true or --multi true if your deployment requires it. If no namespace is specified, the agent installs in the default namespace precisely-ns.

Standalone mode

To install an agent in standalone mode, follow these steps:

  1. On the main navigation menu, select Configuration > Agents.
  2. Select +Create Agent to start the configuration process, then enter the name and description for the agent and choose Create to finalize.
  3. Copy the download agent CLI command by selecting the Copy icon next to it. Paste the command into your terminal to download the CLI package. 
    wget https://cdn.cloud.precisely.com/agent-cli/discli && chmod +x ./discli

  4. Before proceeding with the installation, ensure your system has the necessary resources by running the following command:

     ./discli agent pre-flight

  5. Once the CLI package is downloaded, copy the install agent command directly from the UI. Paste this command into your terminal, and then run the command.

    ./discli agent install --agent-id <id> --regKey <key> --installUrl 
https://agent-management.dis-platform.cloud.precisely.com/agents
  6. Recommended flags for flexible installation : Several flags are available to support flexible, environment-agnostic installations while remaining backward compatible:
    Table 1.
    Flag Description
    --install-mode <standalone> Installs in single-node mode
    --namespace <user-given-namespace> Installs components in user-defined namespace
    --namespace <namespace> Specifies the target namespace for installation. Default: precisely-ns
    -x true or --multi true Enables multi-namespace installation mode.
    --install-modules <list> You can choose to install all services or a subset using a comma-separated list. Examples:
    • --install-modules DI: Installs all Data Integration modules.
    • --install-modules connect-cdc,spark-connector: Installs only the specified modules such as connect-CDC and spark-connector.

Cluster mode

To install an agent in cluster mode, follow these steps:

  1. On the main navigation menu, select Configuration > Agents.
  2. Select +Create Agent to start the configuration process, then enter the name and description for the agent and choose Create to finalize.
  3. Copy the download agent CLI command by selecting the Copy icon next to it. Paste the command into your terminal to download the CLI package. 
    wget https://cdn.cloud.precisely.com/agent-cli/discli && chmod +x ./discli
  4. If your infrastructure is hosted on AWS and the agent is deployed on Amazon Elastic Kubernetes Service (EKS), additional Kubernetes access configuration is required before installation. For more information, see EKS, GKE, and AKS cluster access setup.

  5. Once the CLI package is downloaded, copy the install agent command directly from the UI. Paste this command into your terminal, and then run the command.

    ./discli agent install --agent-id <id> --regKey <key> --installUrl 
https://agent-management.dis-platform.cloud.precisely.com/agents --install-mode cluster --storage-class <storage-class> -n <namespace>

    The command starts the cluster-mode agent installation.

  6. Recommended flags for flexible installation : Several flags are available to support flexible, environment-agnostic installations while remaining backward compatible:
    Table 2.
    Flag Description
    --install-mode <cluster> Deploys to Kubernetes cluster
    --namespace <user-given-namespace> Target Kubernetes namespace
    --install-modules <list> Install selected modules
    --storage-class <storage_class_name> CSI-compliant storage class for dynamic volume provisioning.
    -n <namespace> Specifies the target Kubernetes namespace for installation. Default: precisely-ns
    -x true or --multi true Enables multi-namespace installation mode.
  7. Helm-based installation : For cluster mode, installation can also be done via Helm.

    1. Requirements: Access to a Kubernetes cluster and helm and kubectl installed on the local machine.

    2. Install command: helm install <release-name> <chart> --set <options>

      Helm installation options correspond to the same configuration parameters available in the CLI method.

Post-installation verification

  1. After installation, the agent will be registered with the system using the provided registration key. You can manage and interact with the agent via the CLI or UI.
  2. To check the agent's health status, return to the Manage Agent page. The status will be indicated by the following icons:
    • Action Required: Displayed when the agent has not checked in for more than two minutes.
    • Service Down: Displayed when one or more agent services are unavailable.
    • Not Registered: Displayed when the agent has been added but not yet installed.
    • OK: Displayed when the agent's check-in is up-to-date, and all services are running.
  3. Select the agent name to open the Agent Details page, where a list of installed services appears.

Note:
  • If step 3 fails, ensure that wget is installed. Alternatively, you can use curl -O to download the file. It is recommended to create a user named Precisely on the VM to install and manage the agent, ensuring consistency across all environments.
  • Be sure to copy the full command with actual values from the UI, the example shown above uses placeholders.
  • The Download Agent and Install Agent commands are provided only once. It is essential to document these commands securely for future reference. Store them in a safe location to prevent loss and facilitate access later.

  • If you lose the commands, delete the existing agent and recreate it to generate new installation commands. If the installation has not succeeded, the install command can be reused a limited number of times before the key expires. Once you exceed this limit, use the "re-register" command to reinstall the agent instead of using the original key or command.

    However, if the installation is successful, the key cannot be reused in its original form. In this case, there is no need to keep the original key or command.

Troubleshoot Metrics Server availability

If the Service Down status appears after you install an agent in cluster mode, and the spark-connector service does not start, check the Metrics Server. A Metrics Server pod in 0/1 Running state can prevent Kubernetes resource metrics from becoming available.

  1. Check the Metrics Server pod status:

    sudo /usr/local/bin/k3s kubectl get pods -n kube-system -l k8s-app=metrics-server -o wide

    A healthy pod shows 1/1 Running. If the pod shows 0/1 Running, continue with the following steps.

  2. Check the active firewalld zone and the zone assigned to the Kubernetes CNI interface:

    sudo firewall-cmd --get-active-zones
sudo firewall-cmd --get-zone-of-interface=cni0

    If the CNI interface is not assigned to a zone, verify the existing firewall configuration before adding a rule.

  3. Check whether the firewall already allows traffic to kubelet port 10250:

    sudo firewall-cmd --zone=public --list-ports
sudo firewall-cmd --zone=public --list-rich-rules

  4. Find the pod network CIDR used by the Agent installation. Replace 10.42.0.0/24 in the following command with the actual pod CIDR:

    sudo firewall-cmd --permanent --add-rich-rule='rule family="ipv4" source address="10.42.0.0/24" port protocol="tcp" port="10250" accept'

    This rule allows the Kubernetes pod network to communicate with the kubelet. Do not assume that 10.42.0.0/24 is the pod network for every installation.

  5. Reload firewalld and verify the rule:

    sudo firewall-cmd --reload
sudo firewall-cmd --zone=public --list-rich-rules

  6. Restart the Metrics Server:

    sudo /usr/local/bin/k3s kubectl delete pod -n kube-system -l k8s-app=metrics-server

    Kubernetes recreates the pod. Wait for the new pod to show 1/1 Running.

  7. Verify that the Metrics API is available:

    sudo /usr/local/bin/k3s kubectl get apiservices | grep metrics

    The v1beta1.metrics.k8s.io API service should show status True.

  8. Verify that Kubernetes returns resource metrics:

    sudo /usr/local/bin/k3s kubectl top pods -A

    The command should display CPU and memory metrics.

  9. Verify the Agent services:

    ./discli agent status

    Required active services should show 1/1 Running. Completed logrotate jobs that show 0/1 Succeeded are expected and do not indicate a service failure.