Datacenter & Cloud

A datacenter with 1000+ VMs, 10+ physical clusters and monitoring stacks has one common denominator: everything must stay in sync. Poor synchronisation means:

• Logs cannot be placed in chronological order, so incident response takes 3x longer
• TLS certificate validation rejects packets, causing apps to fail unexpectedly
• Database replication drifts out of sync, creating a data corruption risk

What we typically deploy:

• 2x NTP100-GNSS for stratum-1 redundancy (active-active)
• GMR1000 or GMR5000 as PTP grandmaster wherever sub-millisecond requirements apply
• At scale: PTP grandmaster with OCXO holdover and boundary-clock switches
• Rooftop antenna installation: gas-discharge arrestor, shielded coax and monitoring

Compliance context:

• SOC 2 / ISO 27001: auditable time source required
• GDPR: timestamping of consent logs and data events
• FINRA / MiFID II (financial-sector datacenters): sub-millisecond traceability

Practical point of attention: an NTP server running on virtualised infrastructure is a red flag, VM clocks drift unpredictably. Our NTP100-GNSS is dedicated hardware that removes that dependency.

We do not advise on DCIM or server hardware, that is what HPE, Dell and Cisco are for. The time foundation underneath your datacenter is our specialty.

Which time server do you choose for a data centre?

For most data centre uses (server sync, logging, NTP distribution), a stratum-1 NTP server on dedicated hardware is enough: the NTP100-GNSS, or the NTP100-OSC if you want long holdover on GPS loss. If an application needs microsecond accuracy (distributed databases, PTP fabric), a GMR grandmaster comes into play. Unsure between NTP and PTP? Follow the NTP/PTP time server decision guide.