Campus fabric migration
Moved a live colocation floor from a three-tier design to EVPN-VXLAN spine-leaf with no customer-visible outage, using a phased cutover and a rollback plan tested in the lab first.
Network engineer, data centre operations
I design, build and keep running the networks that sit behind other people's services: spine-leaf fabrics, resilient WAN edges and the facilities they live in.
I have spent over two decade's moving between the network team and the data hall floor, which means I'm as comfortable reading a BGP table as I am checking cooling containment and power redundancy before a cutover.
My focus is uptime that doesn't depend on heroics: documented designs, tested failover, change windows that finish early and monitoring that tells you something is wrong before a customer does.
2013 to now
Lead the network design for a 4 MW London campus. Rolled out an EVPN-VXLAN spine-leaf fabric, cut provisioning time for new customer cross-connects from days to hours, and own the on-call escalation path.
2009 - 2013
Ran day-to-day operations across three facilities. Migrated core routing from static design to OSPF and BGP with BFD, and built the standard rack build and cabling documentation the whole team still uses.
2002 - 2009
Supported enterprise WAN and LAN estates for 60+ customers, handled L2 and L3 escalations, and carried out hands-on installation of switches, firewalls and structured cabling.
BGP, OSPF, IS-IS, EVPN-VXLAN, MPLS, VLANs, STP, LACP, QoS
Spine-leaf design, structured cabling (copper and fibre), power and cooling basics, capacity planning, DCIM
Palo Alto and Fortinet firewalls, ACL design, segmentation, IPsec and SSL VPNs
Python, Ansible, Netmiko, Git, SNMP, NetFlow, Grafana, Zabbix, LibreNMS
Cisco Nexus and Catalyst, Juniper QFX and MX, Arista EOS
Moved a live colocation floor from a three-tier design to EVPN-VXLAN spine-leaf with no customer-visible outage, using a phased cutover and a rollback plan tested in the lab first.
Built a Python and Git pipeline that pulls device configs nightly, diffs them against approved templates and raises tickets on drift. Unplanned config changes fell by roughly 80%.
Designed diverse-path 100G links between two sites with BGP and BFD for sub-second failover, and ran quarterly failover tests with the operations team.
Open to senior network and data centre roles, and to consulting on migrations and resilience reviews.