Health Check step timing out

[hardikk (Hardik Kamdar)]

Hi,

I have an Octopus deployment process where I am using the “Health Check” step in the process. Due to a latest windows patch, it takes some time for my tentacle to get online. In the meantime, the “Health Check” step runs and times out after 1 minute and the deployment fails. I want to increase the health check timeout in the step from 1 min to say 3 mins. Im unable to find any option to customise that. What would be the solution here?

The Octopus version Im using is 3.11.17.

[Alex.Rolley]

HI Hardik,

Thanks for getting in touch!

The timeout in our Health Check step isn’t configurable, unfortunately. What I would recommend in this instance is a script step that has a simple wait timer to allow the machine to restart before running the health check step. A simple powershell sleep should do the trick:

start-sleep -s 120

Let me know if there is anything else that I can assist with,

Regards,
Alex

[hardikk (Hardik Kamdar)]

Hi Alex,

Thank you so much for getting back. I did think about the same thing. However, I won’t be able to do that because of the following:

1. I have the Octopus software installed on the base image that Im using
2. When the instance boots up, it registers itself into Octopus, but the tentacle services are not yet installed on the system
3. If I add a “sleep” command, Octopus does the health check in the meantime on its own and classifies the instance Unhealthy
4. Im using a custom machine policy for the instance that removes the instance if the machine is unavailable for 1 minute
5. Therefore, before my explicit health check comes into play when the services would have started on the system, Octopus removes the machine from the system
6. I cannot change the custom machine policy to a great extent, because those parameters have been configured for a specific reason.

I hope Ive been able to explain the situation much better?

Also, is the timeout in the Health Check configurable in the latest versions of Octopus?

Thanks,
Hardik

[Alex.Rolley]

Hi Hardik,

I’m extremely sorry for the delay in getting back to you on this one! Entirely my fault, I had this reply written up late last week and forgot to push send/

I’ve discussed this with some engineers, and without being able to change the timeout in you custom policy we can’t think of a way around this pickle. We also discussed that even if the health check timeout was configurable (which it isn’t) your custom policy would come along and remove the machine anyway (much as it does with the sleep step).

Sorry for the bad news, but unless you can relax your time out we can’t think of anything that would resolve this issue.

Regards,
Alex