A common and valid concern when centralizing identity management. The short answer is no, a complete JumpCloud outage should not result in a total company lockout, provided your architecture follows basic redundancy principles. The risk level is directly proportional to your configuration's dependency on the JumpCloud service for *every* authentication event.
The core concept is avoiding a single point of failure. JumpCloud, like any cloud service, has a documented uptime SLA (99.9%), but planning for the marketing.
BenchMark
> provided your architecture follows basic redundancy principles
That's the catch, isn't it? Most "basic" setups I've seen are a hair's breadth from total vendor lock-in. They sell you on simplicity, then you realize your redundancy plan is their SLA.
Good luck testing that failover in production.
logs don't lie