So I understand mainframes process data in batches. With millions of transactions per second, how is the frontend getting the most up to date data if the next batch hasn't run?
SnapOverflow
So I understand mainframes process data in batches. With millions of transactions per second, how is the frontend getting the most up to date data if the next batch hasn't run?
In a very simple view: The word 'mainframe' relates to a type of computer hardware architecture, which differs from the Desktop PC design. Mainframes are usually designed for use in large organizations and it also must have an operational system and applications designed accordingly to their purpose. In the last decades the Desktop PC architecture evolution was astonishing, substituting and confining even more the mainframes to very specific uses.
Your question is relevant. Indeed, online transactions must somehow be stored and computed in real time, and they are. Many companies don't need a mainframe for this, nowadays it's usual to find a powerful desktop architecture computer running ERP applications capable of online processing and batch analytical processes. Online transactions occur in real time retrieving and storing data in matter seconds. The batch processes aims different purposes, running usually on a overnight basis in order to produce Analytical data,for business support, reports, data consolidation, etc. If needed, of course mainframes also could do the same.