OPCDA Performance
OPC Server connection
This examination needs 2 OPC Servers. Our OPC Servers is used for the Client-side and another is used for the Server-side.
It is measured reading time of Server-side's Common memory. Communication performance is defined as average time of ten times measurement.
This time means to read the value of a continuous area of the device.
PC specification(Client-side)
OS:
    Windows 7 Professional  32bit OS
CPU:
    Intel(R) Core(TM) i7 CPU 860 @2.80GHz
Memory:3.00GB
OPC Server:
    DeviceXPlorer OPC Server Ver4.2.0.2
OPC Client:
    OPCClient sample VcSampleOpc(C++)
Cable:
    100BASE-TX cross cable connecting directly
OPC Server setting
Protocol Timer Tick:10msec
Timer Tick for Async Update Mode:10msec
Multi Access Mode:Enable
Multi Access for Snyc Asnyc:Disabled
Async Update Mode:Disabled
PC specification(Server-side) that is connected by Client-side PC
OS:
    Windows XP Professional SP3 32bit OS
CPU:Intel(R) 4CPU @2.80GHz
Memory:2.00GB
OPC Server:
    MELSEC OPC Server Ver3.12 build 6
Register:D0-D8191(Simulation mode as common memory)
Communication Performance
≪Local connection≫
(sec)
|
| 1024
| 2048
| 4096
| 8192
| 16K
| 32K
| 64K
| 128K
|
| Average
| 0.001 |
0.002 |
0.004 |
0.006 |
0.015 |
0.031 |
0.048 |
0.092 |
| Max
| 0.001 |
0.003 |
0.004 |
0.008 |
0.017 |
0.033 |
0.064 |
0.119 |
| Min
| 0.001 |
0.001 |
0.004 |
0.003 |
0.008 |
0.027 |
0.024 |
0.059 |
≪Remote connection(DCOM)≫
(sec)
|
| 1024
| 2048
| 4096
| 8192
| 16K
| 32K
| 64K
| 128K
|
| Average
| 0.002 |
0.002 |
0.004 |
0.008 |
0.015 |
0.043 |
0.054 |
0.095 |
| Max
| 0.003 |
0.003 |
0.004 |
0.008 |
0.015 |
0.160 |
0.065 |
0.133 |
| Min
| 0.001 |
0.002 |
0.002 |
0.006 |
0.007 |
0.025 |
0.028 |
0.070 |
Results
In case of continuous Common memory, it is possible to get max 128,000 points per 0.1 sec. And it is neary time between local connection and DCOM connection.
Comment
*Graph displays average time of ten times measurement.(1point = 1word)
*In case of 32K points access , Read 32K points by using 2 devices.( 16K point per 1 device).
*The result only means our examination, it doesn't guarantee performance.
[Jun 2010]
Interface performance
Results of measurement of each interface(eg.sync /async /datachange event etc).
(detail)