[bug] Deal with shared storage sidecar when fail

[luiztauffer]

When initializing a Piece in kubernetes and active shared storage, the pod waits for the confirmation of the shared storage's sidecar container initialization. However, if an error occurs during the initialization of this sidecar, no signal is sent to the base container of the pod, causing the pod to be stuck until the timeout (~15min).

To solve this issue, we should check the initialization of the sidecar and, if there is a failure, send an interruption signal to the base container. Additionally, if possible, it is important to provide some kind of log to the user, informing them of what error occurred.