How it Works - Page 2 of 3
Collating and distributing transactions - Where offices are not permanently connected to each other
Branch update
Periodically each branch retrieves the combined transaction file distributed from Central, checks it, decrypts it and extracts the transactions which it then merges with its own dataset, bringing its data up to date with all the other branches and Central.
Periods for collation and merging
You can choose any periodic interval for collation and merging. The minimum practical interval is probably an hour. However, initially we would recommend daily updates with the process being run automatically overnight. This means that each morning, each branch has a dataset fully updated with all company wide transactions up to close of business the previous day.
Ad hoc and unilateral updating
Apart from the routine operation as set out above, the ERA can deal with ad hoc single or both ways updates from one or more branches at any time. This makes it possible to deal with urgent situations. Take as an example where an immediate position report is required for a job and it is known that staff at two branches have unentered time. The branches could enter their staff time immediately for those jobs and execute an ad hoc data update. Central executes an ad hoc data collection and a merge and produces the report; within minutes of the time being entered at the branches. There is no need for Central to do a corresponding distribution, that can wait until done as routine, although in the example it could be done just to the two affected branches. None of this will upset the normal routine operation and there will be no double counting of the data already sent. The ERA knows at all times the status of all transactions on the system.
Alternatively staff at Central could execute an ad hoc distribution at any time during the day if they wished urgently to distribute newly update data to all or some branches.
Concurrent operation
It should be noted that whilst the branch or Central servers are either collating or updating transactions, routine day to day operation can continue, including data entry and report production. There will only be any perceptible performance degradation if the server is not powerful enough.
