When hyperion is running in a container, the existing run_hyperion.sh will no longer be able to restart hyperion because it will be invoked from outside the container.
As an interim workaround, this can be achieved by scaling down the pods to 0 and then back up to 1 in the kubernetes dashboard.
Longer term we will need a solution, it could be implemented as one of the following:
Implement the currently unimplemented Actions.SHUTDOWN command to shutdown the server - we already implement STOP but this only stops the current plan. Once the existing service instance exits, k8s should then restart it naturally.
hyperion_restart() could invoke a script that manages the kubernetes deployment directly, however it would need to be running with the appropriate privileges so this would be dependent on how the account is set up.
I think option 1 is probably simplest
Acceptance Criteria
Non-containerised deployments of hyperion are still able to restart without requiring any changes
hyperion_restart() on a containerised deployment restarts hyperion
When hyperion is running in a container, the existing
run_hyperion.shwill no longer be able to restart hyperion because it will be invoked from outside the container.As an interim workaround, this can be achieved by scaling down the pods to 0 and then back up to 1 in the kubernetes dashboard.
Longer term we will need a solution, it could be implemented as one of the following:
Actions.SHUTDOWNcommand to shutdown the server - we already implementSTOPbut this only stops the current plan. Once the existing service instance exits, k8s should then restart it naturally.hyperion_restart()could invoke a script that manages the kubernetes deployment directly, however it would need to be running with the appropriate privileges so this would be dependent on how the account is set up.I think option 1 is probably simplest
Acceptance Criteria
hyperion_restart()on a containerised deployment restarts hyperion