We step a step template that can be added at the end of the process, but fires when a deployment goes into guided failure. We can then run a script or trigger a notification to a Slack channel

There is now something that allows an email to be sent if certain events happen, if anyone (like me) came here looking for this functionality it's here:
https://octopus.com/docs/administration/managing-infrastructure/subscriptions

I like the idea of a specific run condition or conditions for handling interrupted deployments. But it could be quite complex working out which steps to run during an interrupted deployment. For instance you might want to send an email and a chat notification by default if a step is interrupted. But their might also be cases where you have a manual intervention step that you don't want the default "deployment interrupted" steps. For instance we do various manual validation steps in production where providing a custom notification to our chat room is more appropriate.

Some suggestions for step conditions - "Deployment interrupted", "Concurrent step interrupted", "Unhandled interruption". Once the deployment has moved past an interruption step I would expect it to be ignored for subsequent interruptions.

Some of our deploys can be started before we leave work, then we check back tomorrow to see the results.

With guide failure mode enabled, when an error occurs, it would mean we won't be able to pick up the issue until the next time we log in again (usually the next day).

If we can add an email alert option with guided failure mode (and even manual intervention step) - that would be very helpful, so we can just start the deploy, and look at any issues raised from email.

Second that again. Introducing some form of notification in guided failure mode would be extremely helpful. Email would be ideal, along with any other notification ideas that would make sense in the given context.

Also, this would be useful to have for the Manual Intervention step.

I second this, we are moving very close to CD now, which means a lot of deployments. It makes sense to do Guided Failure, as we some times see connection drops to the agents. Not having to have a guy actively oogling the deployment progress screen would be an opportunity to do more productive stuff while deploying.

Agree this is going to be quite helpful.. we have quite a few process steps to go through, and a deploy can take up to 6-7 hours to complete.

We have environments that deploys a large set of database backups so have caused the deploy to take a long time.

In an event where a failure occurred and goes into guide failure mode, it becomes an issue if we drift off to do other tasks and don't check until a few hours later, and depending on the failure we might have to start again which means another day almost gone.

We use Octopus with Bamboo, but the idea is still the same. Having Octopus somehow notify us that it is waiting would be a huge help. Perhaps a step with a run condition of "Deployment Interrupted" or something like that where we can set it up to send us an email, or a HipChat message, etc, would work? Thanks!