Timeout for update operation

[jaroslaw-pieszka]

There are many timeouts defined for different steps and operations.
Some hard-wired in the code, some configurable.

For update operation we have timeout set to 40 minutes, which happens to be insufficient e.g. when MachineType is updated.

SRE suggests following approach as for update:

Introduce configuration for this timeout value in the KEB code and helm chart, with default having 120m (2 hours).
SRE would like to specify overrides for this configuration per each KCP environment.`

• reconsider all timeouts involved during updating process (steps etc.)
• take into consideration that some paths of execution are relatively short (administrators) and some take a lot of time (MachineType)
• perhaps we need to prepare consistent approach for all timeouts used by KEB, create separate tickets - this ticket is focused on update operation.

[MarekMichali]

This is connected with #981

[MarekMichali]

In update operation:

• check step has hardcoded 40 min
• check_runtime_resource has a configurable timeout through the APP_PROVISIONER_RUNTIME_RESOURCE_STEP_TIMEOUT key. The key is also used in provisioning operation

Other update steps have no timeout