ManageEngine Endpoint Central – Installation Overview

Know What You Are Installing

Endpoint Central is ManageEngine's endpoint-management platform: one central server, one lightweight agent on every computer you manage, and one web console that acts as the control point for the whole network. From that console an admin handles patching, software deployment, inventory and asset scanning, and remote control — without walking to each machine.

The agents do the work on each PC: they take patch binaries and task instructions from the server, and they check in at user or device logon and on a 90-minute refresh cycle. The server can sync with Active Directory or a workgroup to discover your computers, and the console itself can be opened from the LAN, a remote office, or from home over the internet or VPN — no separate client software is needed for the admin.

Check the Server Hardware First

Before running the installer, match the server machine to the number of computers you will manage — ManageEngine publishes the requirement per endpoint count. For the two starting tiers:

1 to 250 computers: Intel Core i3 (2 core / 4 thread) at 2.4 GHz on a physical machine, or 4 virtual processors on a VM; 4 GB RAM; 10 GB of hard disk space.

251 to 500 computers: the same processor class — Intel Core i3 (2 core / 4 thread) at 2.4 GHz, or 4 virtual processors; 4 GB RAM; 10 GB of disk space.

Larger networks scale up from there in further tiers (501–1000, 1001–3000 and beyond), and bigger or distributed sites can add Distribution Servers. On the agent side each managed PC only needs an Intel Pentium-class processor at 1.0 GHz, 512 MB RAM and 3 GB of disk, and the network card should be at least 1 Gbps. The server runs on a Windows machine and the console opens in Microsoft Edge, Mozilla Firefox, Google Chrome or Zoho Ulaa at 1280 x 1024 resolution or higher.

Choose the Database — PostgreSQL or SQL Server

Endpoint Central stores everything it collects — inventory, patch status, configurations — in a database, and two are supported: PGSQL (PostgreSQL) and MSSQL (Microsoft SQL Server).

For most fresh installations the bundled PostgreSQL is the simplest choice: it installs with the product, needs no separate licence, and is what the overview walks through. Choose Microsoft SQL Server instead if your organisation already runs a managed SQL Server infrastructure — the supported versions are SQL Server 2016, 2017, 2019, 2022 and 2025, and a dedicated SQL host in the small tiers is sized like the product server itself (Intel Core i3 class, 4 GB RAM, 10 GB disk, Standard or Enterprise edition). Decide this before installing, because the wizard asks for it during setup.

Open the Ports Agents, Scans and Remote Control Need

Endpoint Central talks to its agents over fixed ports — if the firewall blocks them, the console installs fine but nothing gets managed. The ports to open on the server are:

8020 — agent-to-server communication and access to the web console. 8383 — secured (HTTPS) communication between agent and server. 8027 — on-demand tasks such as inventory scanning, patch scanning, remote control, remote shutdown and moving agents between remote offices. 8443 — the Remote Control feature over a secured connection. 8444 — the Remote Control feature. 8031 — file transfer in secure mode. 8032 — file transfer. 389 and 636 — communication with the Active Directory domain controller (LDAP and secure LDAP).

On the agent side, ports 135 (remote administration) and 139 & 445 (file and printer sharing) are needed to push-install the agent remotely — ManageEngine notes these may not be required once the agent is installed.

Run the Installer and Open the Web Console

With the hardware checked, the database decided and the ports open, run the Endpoint Central installer on the prepared Windows server and work through the wizard: confirm the installation, choose the database from Step 3, and keep the default ports from Step 4 unless your network forces a change.

When setup completes, the Endpoint Central server runs as a service on that machine. Open a supported browser and go to the server's console address — the server name or IP on port 8020 for the standard console, or port 8383 for the secured connection. Sign in with the default administrator account and set a strong password immediately, before you deploy a single agent.

Verify the Server Is Listening and Reachable

Do not trust a finished wizard — verify the two things that break most often: the port is really listening on the server, and a client machine can really reach it.

From a client machine, open PowerShell and test the console port. Replace YOURSERVER with your server's name or IP address:

Test-NetConnection -ComputerName YOURSERVER -Port 8020

The result you want to see is TcpTestSucceeded : True. Then, on the server itself, open Command Prompt and confirm a process is listening on the same port:

netstat -ano | findstr :8020

If the PowerShell test fails while netstat shows the port listening locally, the block is between the client and the server — go back to Step 4 and re-check the firewall rules for port 8020 before deploying agents.