Difference between Master/Slave and Client/Server

Owned by Rodney Hughes
Last updated: 16 Jun 2015
2 min read

Master/Slave and Client/Server are two different terms with some overlap.

Master/Slave is based on a Master sending an instruction to the Slave, then the Slave carries out that command. 
If there is a command to open a CB the slave will do it. 
If the command is to send back the position of the CB, the slave will then send that information back.
Obviously it can have multiple Masters but it can only carry out the commands of one Master at a time.
i.e. each Master would individually have to ask the Slave what position the CB is in.

So if a protection operation trips the CB, the only way a Master/Slave would know about it is if the Master, by chance, asks the Slave what is the CB position whilst the CB is still open.


If in the meantime there has been an autoreclose before the Master polls for the status, the next request from the Master will result in the Slave simply sending back that the CB is (apparently still) closed  - the fact of the CB opening and reclosing is missed.  (sad)   The Master would have to poll other information to eventually detect that a protection trip occurred followed by the reclosure.

Client/Server does everything Master/Slave does as above PLUS the Server can automatically send information to the Clients
So the Server can be configured to send a Report to the Client(s) every time the CB position changes - noting the Report can be configured to be sent to multiple Clients.

So in this protection trip followed by autoreclose example, if it is Client/Server, the Server would Report the CB opening (as a result of the protection operation) automatically to the ALL the configured Client(s) who are supposed to know about it, and also Report the CB closing again to the Client(s).
The only possibility of the Clients not knowing all operations is if the Report doesn't arrive at the Client(s)
Hence we have two types of Client/Server Reports - Unbuffered and Buffered.

Client/Server Unbuffered Reports means the Server just "throws" the Report onto the comms network without knowing if the message could or does get to the required Clients - if the comms is down anywhere the message would be lost

Client/Server Buffered Reports means the Server effectively verifies the attempted message does in fact get through to the Client.  If it doesn't get verified as reaching the Client,  the Server puts the message in a buffer and keeps retrying to send it until it is verified  (or it is overwritten as a buffer full).