No package for the action and machine was acquired

We are running Octopus Deploy 3.4.5 and use Triggers to initiate deployment to new tentacles in AWS. We use Artifactory Professional 4.13.2 rev 40285.

We periodically receive the following error with trigger initiated deployments:

The step failed: Activities failed with errors XYZ.Web on i-036aea16: ‘No package for the action ‘04e16b78-65c8-4f46-b7f1-551ed8e398ff’ and machine ‘Machines-1016’ was acquired.’

When this happens it happens for all packages, and all machines that are involved in that deployment.

I’ve searched and found this (Deployments can fail if a manual step happens after package acquisition · Issue #1782 · OctopusDeploy/Issues · GitHub) but it looks like this issue was resolved with version 3.0.7.

Can anyone help me to diagnose this issue?
Thanks!

Hi Luke,

Thanks for getting in touch.

Would you mind sending a complete task log for a deployment where you are seeing that error?

We made a lot of package acquisition changes in 3.4.11 so updating to latest may help.

Cheers,
Ahane

Yes, don’t want to post here though. What is your email address Shane?

Hi Luke,

Send it to support@octopus.com and I’ll pick it up.

Cheers,
Shane

Sent, thanks!

Hey Shane. I’m taking over this from Luke. Can you please send me your contact details so I can get those logs to you?

Hi John,

Luke sent me the unsuccessful logs, would you mind sending logs when an auto-deploy happened and it was successful?

Please send them to support@octopus.com.

Cheers,
Shane

Thanks Shane. Attached is what looks like a fully successful deploy.

FYI - have raised a seperate ticket just now about a problem with auto
deploy skipping a step that it shouldn’t. Probably not related but perhaps
something you can also help with?

Cheers

John

ServerTasks-15057.log.txt.7z (64 KB)

Hi Shane. Had any luck with this? Or need any further info?

Hi John,

I’ve been able to reproduce both of the issues you are seeing, both bugs we will fix shortly.

It looks like automatic deployments are running steps that were manually skipped during the initial deployment which is causing the package acquisition error.

The steps that aren’t run I believe is because the steps are scoped to multiple roles.

Cheers
Shane

Hi guys,

We’ve made a bunch of fixes for automatic deployments in 3.4.15 to address the issues you’ve been experiencing. Please let me know how you find it.

Cheers,
Shane