6/19/2010 12:58 PM | |
Posts: 1 Rating:
|
Hello~
First of all, I ‘ll explain to my system.
Primary rack configured like below Backup rack configured lke below
From Mr.Kim AttachmentCPU Cycle time.zip (408 Downloads) |
Last edited by: shineKim at: 6/20/2010 6:37 AM1. CP cards firmware version was V2.7.3.and i modified acticle. |
|
6/20/2010 7:53 PM | |
Joined: 12/22/2006 Last visit: 1/25/2022 Posts: 1742 Rating:
|
Hello, I don't believe that the cycle time is the main cause of what you see. In this manual: 400h system manual In chapter 11.7, they explain more in detail the following points: 1)compared to stand-alone the communication in redundant mode is significantly lower 2) there is a linear increase of throughput until stauration is reached. 3) fault tolerant mode requires synchronization to maintain synchronuous operation. Data arriving in one CPU must be transferred via the sync link in the memory of the other CPU. With the amount of communication you have listed, I am afraid that this is your main problem. And I guess that you have several Modbus commands over a single connection. What means a lot of communication jobs. 4) some points will influence the performance. One of them is the number of connections. And you have reached nearly the maximum possible because you have 46 connections one-sided + 3 DCS + 3 DCS What can help? - reduce the rate of communication jobs per second as far as possible - group data to reduce the number of jobs (each request requires time for execution and synchronisation) - call blocks sequentially and stepped in order to balance distribution of load - if possible, make the communication event driven and by-pass the block call command if you do not want to transfer data I hope this helps |
Best regards |
|
This contribution was helpful to
3 thankful Users |
Follow us on