FileConnect Enterprise v3.1 Setup

  • Updated

FCE 3.1 is the only version supported for new installations. It installs directly onto your host — no virtual machine (OVA) is required or supplied. Existing FCE 3.0 (OVA) customers should follow the FCE 3.0 → 3.1 Upgrade guide instead.

 

This guide explains how to install, configure, and activate FileConnect Enterprise v3.1 (FCE 3.1) in a production environment. It covers prerequisites, connector configuration, installation, licensing, and verification.

1. Before you begin

Confirm your host meets the FileConnect Enterprise Host Requirements. At a minimum you need:

  • A supported host OS: RHEL 9 or 10, or CentOS 9 or 10 
  • 4 vCPU, 32 GB RAM, and disk sized at ≥200 GB per TB of managed data
  • Root or sudo access to the host
  • Network access from the host to the file shares FCE will manage, and to the required external RecordPoint endpoints

FCE installs most of its own dependencies (EPEL, Python 3.11, Ansible, JDK 21, and a single-node Elasticsearch 8). For larger deployments, a 2–3 node Elasticsearch cluster is recommended - contact RecordPoint for the FCE Advanced Deployment guide.

 

2. Default user accounts

 

System AppUsername PasswordPurpose
Diskover app admin / darkdata Diskover web application admin (change on first login)
Elasticsearch elastic / (generated) Password stored in /root/.config/elastic.txt at install; after FCE config, retrieve with the helper script below
RabbitMQ Management diskover / (you set this) Management UI at http://localhost:15672

Retrieve the Elasticsearch password after FCE is installed:

bash /opt/recordpoint-connector-IndexBuilder/Deployment/get-appsettings-elastic-pass.sh

3. Get your access token

FCE 3.1 requires an access token to install and update the Diskover components. Contact RecordPoint to obtain an access token exclusive to your account and store it securely — you'll use it whenever you deploy or update FCE.

4. Register the Azure application

  1. Sign in to the Azure Portal with an admin account.
  2. Go to Entra ID → App registrations → New registration.
  3. Set Name to FileConnect Enterprise and Supported account types to "Accounts in this organizational directory only," then Register.
  4. On the Overview page, copy the Application (client) ID.
  5. Go to Certificates & secrets → New client secret. Add a description (e.g. "fce secret") and a 12–24 month expiry.
  6. Copy the secret Value (not the Secret ID) and store it securely — it cannot be retrieved later.

5. Configure the connector in RecordPoint Data Governance

  1. Sign in to your RecordPoint Data Governance region (e.g. https://usw.records365.com).
  2. Click the gear icon → Connectors → + Add Connector.
  3. Add the FileConnect Enterprise (v3) connector.
  4. Enter the Azure Application (client) ID.
  5. Set Aggregation Level to Lowest Level Folder.
  6. Optionally rename the connector, then Save and Enable.

6. Download the installation files

From the connector configuration page, download:

File Description
Configuration Settings (JSON) Connector configuration; imported later in the FCE UI
FCE Installer (ZIP) The latest FCE installation files (includes install-Diskover.sh and install-FCE.sh)
Copy the ZIP to the host and unzip it. All commands below run from the unzipped installer directory as a user with sudo privileges.
 

7. Prepare prerequisites

Install EPEL (required on RHEL-compatible systems) per the official Fedora EPEL instructions.

Set the RabbitMQ password used by FCE:
 

tar -xzf ansible.tgz
cd inventory/group_vars/

Open all.yml, find the RabbitMQ password setting (near line 51), and replace the placeholder XXX with a secure password. Return to the main installer directory.

8. Install Diskover and Elasticsearch

install-Diskover.sh installs the Diskover component, RabbitMQ, and a single-node Elasticsearch on the host. The only required parameter is the access token (run with -h for all options):

 
sudo ./install-Diskover.sh --access-token "<name>:<token>"

Reload components after it completes:

 
sudo nginx -s reload
sudo systemctl daemon-reload

Back up your deploy config: copy /inventory/hosts.yml and /inventory/group_vars/hosts.yml somewhere safe — you'll need them for future Diskover upgrades.

9. Prevent unintended package upgrades

To stop a routine dnf update from upgrading critical components, add an exclusion list:

 
sudo gedit /etc/yum.conf

Add:

 
exclude=nginx* php* diskover* rabbitmq* kibana* elasticsearch*

When a major upgrade is due, comment this line out to allow the packages to update.

10. Configure Diskover

  1. Open the Diskover web UI at http://localhost:8000/diskover_admin/config/ and log in with the default credentials.
  2. Work through each configuration section. Where a Test button is available, test the connection, then Save and continue.
  3. Elasticsearch: confirm the connection settings for both diskoverd and diskover-web, then test.
  4. License: copy the displayed hardware ID and share it with RecordPoint. RecordPoint returns the Diskover and Diskover-Web license values — paste them in and save.
  5. Message Queue: for the diskover user, enter the RabbitMQ password you set in step 7.
  6. Start the services:
 
sudo systemctl start diskoverd celery
sudo systemctl restart diskoverd celery diskover-admin
  1. Validate access: open http://localhost:8000, log in as admin, and change the default password. "No completed indices found" is expected at this stage.

Store the Elasticsearch password for later:

 
sudo cat /root/.config/elastic.txt

11. Install FCE

Run the FCE installer (--verbose shows live output, which is also written to log files):

 
sudo ./install-FCE.sh --verbose

Optional password parameters are supported (--interactive, --elastic-password, --rabbitmq-password) — contact RecordPoint for the Security Hardening FCEv3 guide.

Verify all services are running:

 
cd /opt/recordpoint-connector-IndexBuilder/Deployment
sudo chmod a+x ./fcectl.sh
./fcectl.sh --status

All services should show running. (The tika-http and tika-http-ocr services may show as disabled — this is expected.)

Enable the Diskover message queue. Edit /etc/rabbitmq/rabbitmq.conf (create it if missing) and add:

deprecated_features.permit.transient_nonexcl_queues = true

Then restart RabbitMQ:

 
sudo systemctl restart rabbitmq-server

Confirm the FCE UI loads at http://localhost:8080.

12. Connect FCE to RecordPoint Data Governance

  1. In the FCE UI, open Settings.
  2. Enable options as needed:
    • Azure Logging → Enable Azure Monitor (sends logs to RecordPoint App Insights)
    • Document Cracking → Extract and Store File Content, or Extract Content Metadata Only (Author) for the Author field without full content extraction
    • Crack Image Types and Enable OCR (optional; require content extraction)
  3. Click Import JSON file and select the Configuration Settings JSON you downloaded.
  4. Paste the Azure App Registration Client Secret into the Client Secret box.
  5. Click Test Settings to validate, then Save.

13. Verify and get started

Updating FCE later

To move to a newer 3.1 build:

  1. Download the latest installer package from the connector configuration page and unzip it on the host.
  2. As a precaution, stop the FCE service:
 
   /opt/recordpoint-connector-IndexBuilder/Deployment/fcectl.sh --stop --fce
  1. Re-run install-FCE.sh to apply and configure the latest version.
  2. Confirm the version on the Settings page at <fcehost>:8080.

Was this article helpful?

0 out of 0 found this helpful

Have more questions? Submit a request