04 · Infrastructure operations · Current role
Running the infrastructure for a group of about 11 companies
Day-to-day responsibility for the servers, virtualisation, network, security, CCTV and end-user IT that a group of about 11 related companies runs on. This is ongoing operations work, not a single project.
The problem
About 11 related companies share one set of infrastructure: servers, network, firewall, Wi-Fi, CCTV and the ERP. When something fails, the business stops, whether it's a disk, a VPN tunnel or a printer at the front desk.
The work runs from the server room down to individual users, so the useful skill is following a problem across layers instead of handing it off.
What I'm responsible for
- Dell PowerEdge R740 / R750 servers: replacing and reinstalling SSDs and HDDs, reinstalling operating systems, running the VMware ESXi hosts and watching their health.
- Windows Server and Ubuntu Linux systems, including the ERP servers.
- Check Point firewall, Remote Access VPN, NAT and routing.
- Grandstream GWN Wi-Fi and switches managed through GWN.Cloud, plus printers and IP telephony.
- NVR camera systems across sites: recording, storage and remote viewing.
- End-user support: accounts, email, Office, PC rebuilds and hardware swaps.
Architecture
A simplified view of the layers I'm responsible for.
- Users & sites
- Office PCs and laptops
- Printers · IP phones
- CCTV cameras
- Edge & remote access
- Check Point security gateway
- Firewall policy · NAT
- Remote Access VPN
- Network
- Switching
- Grandstream GWN Wi-Fi via GWN.Cloud
- Compute & workloads
- Dell PowerEdge R740 / R750
- VMware ESXi hosts
- Windows Server and Ubuntu VMs
- NVR recording
Engineering decisions
Troubleshoot by layer
I start at what the user sees and work down through the application, database, operating system, network and firewall until the cause is found. Most problems that look like application bugs end up being somewhere else.
Fix the cause, then write it down
Restoring service comes first, but a problem isn't closed until I know why it happened and have done what I can to stop it coming back.
Separate test from production
New systems like the Odoo 19 migration run on separate test servers and databases instead of being tried out on what the business depends on.
Technical challenges
Hardware failures on production servers
Replacing failed SSDs and HDDs and reinstalling operating systems on PowerEdge servers, while keeping downtime for the companies as short as possible.
Remote access across companies
Keeping Check Point VPN access working for staff who need to reach internal systems from outside the office.
Breadth
In one day the work can cover an ESXi host, a firewall rule, an NVR that stopped recording and a laptop that won't boot. Prioritising by business impact matters as much as technical depth.
Current status
Ongoing
- Server, virtualisation and storage operations
- Firewall, VPN and network operations
- CCTV / NVR and end-user support
Outcomes so far
- Servers, network and CCTV kept running through hardware failures by replacing disks and rebuilding systems.
- Built gkllc.mn, the company website whose job listings come straight from the ERP, with CV uploads that land in Odoo HR.
This is internal infrastructure. I don't publish uptime, incident or cost figures, because I can't share the data behind them.