Skip to main content
Most teams have heard “it will just be more noise” about every alerting tool. Shadow mode lets you check that claim on your own data before anything about your current setup changes.

​
How it works

Ordyn posts alerts only to the Slack channels you turn on. In shadow mode, you turn on one new channel, such as #struct-alerts, and leave your existing channels exactly as they are. Nothing is rerouted and nothing is turned off. For a week, both channels run side by side. Your current channel shows what you get today. The Ordyn channel shows what the same period looks like when every alert is investigated and graded first.

​
Set it up

1

Finish the basics

Connect GitHub, Slack, and at least one observability source. See the Quickstart.
2

Route Ordyn to a new channel

In Detect setup, or later in Triage, create a new Slack channel for Ordyn and describe in plain language what belongs there. Leave your existing alert channels off.
3

Turn on Auto Monitors

Ordyn starts working on the next pull request. Monitors are created as changes deploy.
To have Ordyn triage the alerts you already get, turn on Auto-investigate for an existing alert channel in Triage. Ordyn replies in the thread of each alert with its findings. This adds to the channel and does not remove anything.

​
What to compare

After a week, look at both channels and ask:
  • How many messages did each channel post? How many of them mattered?
  • When something real happened, which channel said so first, and did it name the cause?
  • Were the root causes right? Every Ordyn claim links to its evidence, so you can check.

​
Graduating

When you are ready, route Ordyn’s alerts to the channel your team actually reads, and tune the sensitivity to taste. Existing monitors keep working. Ordyn only adds coverage and triage on top.