Fix startup cycle logic

[jarongrigat]

This basically calculates how long a pipeline will throw errors on startup because of unfilled buffers. During this time pipelines are silent and logs are only on DEBUG level since the errors are expected. There were a few things wrong with the implementation:

• The number of startup cycles only depends on the longest buffer of a 'strict' BufferNode.
• There was a reference error where n should be node as the loop says
• Since self.cycles starts counting at 0 the comparison to self.startup_cycles should be < no <=
• Also, we write startup cycles to the database now so we can use it to improve the status indicators on Doberview