I’m about to break one monolithic process into two.
Process A will produce one output excel file (result1.xlsx) that contains transactions for further processing by Process B.
Thus, Process A will on occasion write to result1.xlsx. Process B will on occasion read that file. Thus, there is a chance for conflict. I assume I could use an try-catch block. If Process A cannot write to the file, sleep for a little time, try again. Vice versa for Process B.
It’s not ideal perhaps. But this monolithic process have two web applications that it interface against. Stability is good but sometimes crashes will happen. We do use a queue but because of the crashes It’s been necessary to clear the queue before restart. Granted, we do not have the best implementation use case of the queue.
Any thoughts on this? Any other ideas? Looking for insight, inspiration, what would you do? How do the pro’s do it when they share information between two robots.