2026-q3-infrastructure-review.pptx

Infrastructure review: Q3 2026
Acme IT team · quarterly review, 6 October 2026
Availability
Public website (web1, web2): available 99.97 % of the time
Primary database (db1): 99.99 %, one planned restart
Incidents: no P1, two P2 (VPN certificate, full log partition)
MTTR: 1 h 40 min (Q2: 2 h 15 min)
What we delivered
Prometheus pilot on three servers, next to Zabbix
Automatic renewal of the internal certificates
Four laptops delivered and prepared for the September arrivals
Restore test of db1 from the night's backup: 38 minutes
Risks
pve1 is our only hypervisor: no failover if it stops
db1 disk usage at 78 % (alert threshold 85 %)
Firewall page overdue for review in the wiki
Q4 priorities
Decide on the Ceph project (owner: Hugo Bernard)
Prometheus in production for all servers (owner: Inès Petit)
Quarterly restore test, documented (owner: Malik Robert)
Next review: January 2027 (owner: Claire Martin)