What is difference between 2oo3 and 1oo2 on Aspect Server?
I have upgrade project from 800xA 4.1 to 800xA 6.0.3.
However, end-user ask us to reduce number of server.
exsting system has separated one Aspect server(2oo3), two Connectivity server(1oo2) and Domain server(1oo2).
If new system has combined Aspect server(1oo2) with Domain server and two Connectivity server(1oo2), Is this OKAY?
It has 26 advant contorllers and 15 client nodes.
I tried to find relevant informaion about tag/client limit in manual(3BSE041434-600_E_en_System_800xA_6.0.3_System_Guide_Technical_Data_and_C
There was no information.
Voted best answer
First: 2oo3 can accomodate more clients
Second: 2oo3 increases redundancy*
*) A 2oo3 system that is reduced to one server will enter Read Only Mode. A network disconnection will have same effect on a singled out server.
A 1oo2 system has no Read Only Mode. A singled out server will remain writeable. This must be taken into account - when 1oo2 servers rejoin, primary WILL OVERWRITE secondary. First provider in service group = primary. Any engineering work, etc made in secondary will go lost if primary version is more recent. This behavior can be disabled on Aspect Directory service (it is called 1oo2 Automatic Recovery). If disabled, secondary will enter Error State in case primary version is different. Maintenance staff then have to decide which server to cold reset.
Yes we can Combine aspect server with Connectivity Server.
but as you said that you have Number of controlle in Plant So, it is batter to seperate Connectivity Server and Aspect Server.
and difference between both is,
2oo3, 2 out of three. If two server in sytem getting down your Plant will work on one Server.
1oo2, 1 out of two. If one server in sytem getting down then your Plant will work on one Server.
Why not just virtualize everything ?
Having separate virtual servers for separate functions makes software and system management and server redundancy much easier. Running all those virtual servers on only 2 physical machines makes maintenance and hardware fault finding easier. Deploying new virtual servers is incredibly easy compared with building new physical machines. Done properly it only takes 5 to 10 minutes to deploy a new server after the first one has been created.
2oo3 redundancy allows your aspect servers to "vote" on whether one server is bad or not. The majority decides who is correct and whether one is wrong. In theory this gives greater redundancy.
In practice, the 800xA system is generally reliable enough that 1oo2 is sufficient for most applications. Of course, its still important to properly backup and maintain your xA system to keep that reliability.
Virtualizing 800xA is an excellent way to reduce the required number of servers and I HIGHLY recommend it. However, it is a different skillset from maintaining separate servers. In addition to windows management skills you also require VMWare management skills. Training for both you AND your end user is important.