Deploy stuck at "Acquire packages"

Legacy Uncategorised
1

Hi,

During a recent deploy we had a tentacle stuck at the “Acquire packages”.

Looking at the log the process was stuck at “Delta compression is enabled for package transfers from the Octopus Server to deployment targets”.
The step was using the Builtin Nuget feed on the Octopus Server.

A restart of the tentacle service fixed this but is this a known issue/bug?

2

Hi John,

Thanks for getting in touch! I’m not aware of a similar issue with package acquisition hanging for deployments. I hope you don’t mind providing some more context for me in case it helps?

  1. Does this happen often?
  2. Once rebooting the Tentacle, does it happen reliably/frequently afterwards?
  3. How large are the packages you are sending?
  4. How reliable is the connection to the Tentacle from the Octopus Server?
  5. How fast is the connection to the Tentacle from the Octopus Server?
  6. How long does the acquisition phase take on rough average?

If the problem is recurring you could try using an external NuGet feed, and on the Package Step, choose the option to download directly on the Tentacle. This pattern can help take load from the Octopus Server because each Tentacle can just download its own copy of the packages directly from the feed. Some customers find this helps them a lot, but as with all of these choices, it does come with some tradeoffs. This pattern helps when you have a LOT of Tentacles, or if you can provide a CDN local to the Tentacles for the packages to replicate into.

Hope that helps!
Mike

3

Hi Michael,

  1. This was the only time we have encountered this.
  2. After tentacle restart we have not seen this issue again.
  3. Total package size was 20MB.
  4. We use a proxy connection for this tentacle; there have been some connection issues in the past.
  5. A 20MB package normally takes 3 seconds to transfer to the tentacles
  6. The acquire phase normally takes 3 to 5 seconds.
4

Hi John,

Thanks for getting back to me! It sounds like everything should work without any problems, so it would be really nice for me to investigate further and understand if there’s a bug waiting to be discovered.

Would you mind sending through the raw task log and the Octopus Server and Tentacle logs for the same time?

Feel free to make this thread private so you can safely attach the log files.

Hope that helps!
Mike

5

We ran into this same issues yesterday. It may have been trigger by a deployment being queued on the tentacle (behind other deployments) and someone losing patience and canceling it. We attempted to create a new release with a new nuget package. That did not help. It was also resolved by restarting the tentacle service. Tentacle is on 3.5.2

6

Hi!

I’m sorry, I hadn’t noticed your post until today! We have found and fixed a few issues where certain tasks would hang silently with no way to recover them - sounds like it could be similar to the issue you’ve reported.

I would recommend updating to the latest version of Octopus where we’ve fixed all of the known issues related to this behaviour.

Hope that helps!
Mike