Communication Problem

[Aamir 0]

System configuration 

We are facing two main problems, our system configuration is shown below

Master PLC: BMEP582020
87 nos. Slave PLC: TM218

Ethernet is our mode of communication.
IP address of slave  PLC are from 192.168.100.1 to 192.168.100.87
IP address of Master PLC, Rack-1 and Rack-2 are 192.168.100.101, 192.168.100.102 and 192.168.100.103 respectively.

Problems stated bellow
1) After power cycle, main PLC takes around 45 min to read data from 87 slave PLCs.

2) Slave PLCs randomly fails (Status led MS shows constant RED), minimum 2 slave PLC shows such indications.

In the architecture the connection between Master PLC and slave PLC group is through Fiber Optic cable.

In our system the groups are divided as follow
From IP: 192.168.100.1 to 5 are in group-1
6-10 are in group-2
Similarly till group-13
And Group-14 contain 4 PLC and junction box is after 3rd PLC and IPs are from 192.168.100.66 to 69

Group-15 also contain 4 PLC but the junction box is after 4th PLC and IPs are from 192.168.100.70 to 73

Group-16 & group-17 contain 7 plcs and the junction box is after 7th plc and IPs  for group-16 are from 192.168.100.74 to 192.168.100.80 and for group-17 are from 192.168.100.81 to 87.

The junction box contain Ethernet switch and optical fiber converter.

All the groups are connected to Master PLC through optical fiber, group-1 to 15 are on one side and group-16 and 17 are on the other side of Master PLC.

[Scadadoctor 0]

Using Modbus Analyzer  from www.parijat.com product, try to find what happens to MB TCP messages from the slaves. Also, it should not take 45 minutes, unless your data is 1000s of registers & is very badly scattered. Modbus Analyzer may also help you find why the system is taking so long.