Live Webinar
It waits for something to break, then breaks everything else on the way over.
If you have watched a VPN drop, a call die or a dispatcher lose a radio channel because a link switched over, you already know. In 45 minutes we take a failover event apart, stage by stage, and show what replaces it.
FeaturingContrivian Lighthouse™ and Contrivian NorthStar™
Registration opens in Microsoft Teams. Can’t make it live? Register and we’ll send the recording.
$900K+
Cost of one hour of downtime for a Global 2000 company
43%
Of downtime incidents start in the network or IT environment, the largest single cause
47%
Of tech leaders say customers are often the first to notice an outage
57%
Of operators say their last major outage cost more than $100,000. One in five topped $1 million.
The Problem
01
The link degrades long before failover decides it’s dead. Calls break up, video freezes, and no alarm fires.
Gray failures often never trigger failover at all.
In 2026, 47% of tech leaders said customers are often the first to detect downtime. (Splunk)
02
The backup brings a new public IP. Every open session, VPN tunnel, voice call and radio-over-IP channel drops.
When the primary comes back, it happens again.
Anatomy of a Failover Event
A remote site loses its fiber and has cellular or satellite as backup. Timings vary by vendor. The sequence doesn’t.
Stage 01
Link still “up,” dropping packets. Calls break up. No alarm.
Stage 02
Probes miss, hold timers count down. Dead air.
03 | Outage
Public IP changes. Every open session drops.
Stage 04
Tunnels rebuild. Users log back in. Some systems need a manual restart.
Stage 05
Everything on a link sized as a spare. Uploads queue.
06 | Outage
Primary returns, IP changes again. A second outage.
The fiber cut caused one problem. The failover design caused four more.
Why Adding Links Makes It Worse
687
Internet outages worldwide in a single week this September, 331 of them at ISPs. The links themselves aren’t getting more reliable. (Cisco ThousandEyes)
01
Hold timers turn a one-second problem into many seconds of dead air.
02
New link, new public IP. Everything tied to the old address breaks.
03
LEO hands off every 30 to 90 seconds. Cellular swings minute to minute. Failover reads normal behavior as failure.
04
It carries nothing until the worst moment. Expired SIMs and lapsed contracts show up during the outage.
05
Going back to the primary is a second outage. Many teams just stay on the backup for days.
What Replaces It | Contrivian Lighthouse™
Contrivian Lighthouse bonds fiber, broadband, 5G, LTE and LEO into one connection. It measures every path every second and steers each application to the best one.
Everything sits behind one static IP that Contrivian owns, so a link can drop without anyone noticing.
Contrivian NorthStar shows the failed link to our NOC, and we repair it while you keep working.
| Category | Failover | Contrivian Active-Active |
|---|---|---|
| Links | One works, the rest wait | Every link carries traffic |
| Trigger | After a link is judged dead | Steers every second on loss, latency and jitter |
| Public IP | Changes with each link | One static IP, independent of any carrier |
| VPN, Voice, RoIP | Tear down and rebuild | Stay up through link loss and satellite handoffs |
| Spare Link Health | Unknown until needed | Proven every second under load |
| Failback | A second outage | Doesn’t exist |
Take It Home
Five tests to run at one of your sites with production apps open. Do them on your own, or with a Contrivian engineer.
Any dropped session means your network depends on failover.
01
Pull the primary during a live video call, VPN session and file upload. Time the recovery.
02
Check the public IP before and after.
03
Degrade it without killing it: 5% packet loss or 300 ms of latency. Does traffic move?
04
Flap it: drop and restore three times in two minutes. Count the outages users would see.
05
Audit the backup: when did it last carry production traffic?
Who You’ll Hear From
Principal Technologist
Principal Technologist at Contrivian. Co-founded Dyn and ran infrastructure at Fastly, and knows more about your network than you know you don’t know.
CEO, Founder and Chief Plug‑Puller
Has a two-word opinion of failover we can’t print here. Tests his connectivity daily in the redwoods.
CMO and Moderator
Keeps Tom and Grant on time. Mostly.
Gartner Eye on Innovation Award
Contrivian Lighthouse™
Contrivian NorthStar™
Contrivian Constellation™
Next Step
45 minutes. No fluff.
One test you can run on Monday.