When deploying a release to multiple environments at the same time, some times the deploy is stuck when updating a IIS Website.
When it fails, it occurs to at least two environments, which may suggest it is a race condition. I cannot reproduce the error reliably, but if I deploy to 200+ environments it will happen at least a couple of times.
Thanks for all that info. As you might know already, we will be adding multi-tenancy support in 3.4, which will make this scenario a lot easier. At the time being (3.3 and below) the best approach is the one you are taking already (Env per customer).
Are the Targets (tentacles) on each Environment unique, or are some of them machines that are running multiple Tentacle instances? I’m interested in knowing if two tasks(deployment steps) are being executed against IIS on the same machine at the same time
Were you able to find a pattern on which set of Environments/Targes get “stuck”?
Could you share us a 2 deployment logs please: One of a successful deployment and another of one that got stuck. If possible try to send logs of 2 environments that were triggered at the same time.
Thank you for your patience. Sorry I’m late on getting back to you.
The tentacles are not unique on each Environment. As we have a lot of environments of each machine they also share the tentacle. However, this also means, as your asking that two deployments step in fact are executed against the same IIS at the same time.
Dalmiro is on leave for a few more days so I am assisting with this ticket.
Would you mind also providing the Tentacle logs from around the time of the stuck deployment.
They can be found in C:\Octopus\Logs on standard installs.
Thanks for the logs and for waiting for my return!
Is this “kaai” site being deployed at the same time on the same machine with the same name on IIS? Because if that is the case, I don’t think IIS will be able to handle it.