Free tool

How much downtime is built into your design?

Every part a request passes through has to be up at the same time. Stack your parts, give them copies and zones, and watch the expected downtime change.

Add a part

Your setup

%

Your assumption for a full zone or data center outage.

$

Expected availability

-

Downtime a year

-

Downtime a month

-

Cost a year

-

    Starting availability figures are the usual published cloud SLAs: 99.5% for a single instance, 99.99% for a load balancer, 99.9% for object storage and queues, 100% for managed DNS. Change any of them to match your providers.

    Want it fixed?

    Design out the single points of failure

    An architecture session walks through your real setup, finds the parts that take everything down with them, and gives you the cheapest order to fix them.

    Renews monthly. Cancel any time.
    Send us the findings