Reliability · 99.999% target
Reliability · 99.999% target Reliability
& uptime.
A phone system that drops customer calls is worth less than nothing. DialPhone targets 99.999% uptime across every product, voice, video meetings, SMS, contact center, backed by service credits when we miss.
Uptime is the single metric that matters most for a business phone system. A web app that drops one request in a thousand is usable; a phone line that drops one customer call in a thousand is a contract risk. That is why most business communications vendors quote 99.999% as a target, and why the gap between vendors who quote it and vendors who actually hit it shows up in real public incident history rather than in marketing copy.
Five-nines means a maximum of 5 minutes 15 seconds of unplanned downtime per year, measured against a clear service definition (which voice paths, which regions, which channels). Beyond that window, service credits apply to the next invoice automatically, no support ticket required. The SLA schedule on the customer agreement describes the exact credit math for each tier.
Three architectural decisions drive that uptime number. Active-active topology: every region runs a live copy of every service; failover is automatic, not manual, and there are no cold standby regions waiting to warm up. Carrier-diverse PSTN peering: voice traffic exits through multiple Tier-1 carriers, so a single carrier event never takes calls offline. Independent region control planes: an incident in one region cannot cascade. The architecture grid below describes each layer with verifiable numbers.
Numbers
Service-level commitments
- Target uptime
- 99.999%
- RTO
- <15 min
- RPO
- <5 min
- Active regions
- 4 US + 2 EU
- Service credits
- Yes
- Status page
- Public
Architecture
How we hit five nines
- Active-active across 4 US regions and 2 EU regions, no cold failovers
- Carrier-diverse PSTN peering, no single-carrier dependency
- Geo-replicated databases with synchronous writes
- Load-balanced, stateless application tier, rolling deploys with zero downtime
- Automated capacity scaling for call and meeting spikes
- Independent region control planes, one region's incident doesn't cascade
Stay informed
Public status page
Subscribe to the public status page for real-time incident updates and maintenance notices. RSS, email, SMS, Slack, and webhook channels all supported.
Reliability FAQ
What does 99.999% uptime actually mean?
5.26 minutes of allowed downtime per year. Beyond that, customers receive service credits per the SLA schedule. Higher than typical SaaS (99.9% = 8.76 hrs/year) because phone downtime has direct business impact.
What architecture supports 99.999%?
Active-active across 4 US regions (Virginia, Oregon, Texas, Ohio) with automatic failover. Carrier-diverse PSTN peering eliminates single-carrier failures. All critical databases geo-replicated with synchronous writes to 2+ regions.
Where is the status page?
DialPhone status & reliability page, real-time and historical. Subscribe via RSS, email, SMS, Slack, or webhook.
How do I claim service credits?
For qualifying SLA breaches, credits apply automatically to the next invoice. For disputes or questions, contact support@dialphone.com within 60 days of the incident.
How often do you do chaos / game-day testing?
Quarterly, full region failover drills. Monthly, component-level chaos experiments in staging. Findings documented and remediated against published timelines.