IO Server Processes Failing with Dual NIC

CitectSCADA 2018 R2 running on on a server with 2 NIC's (SCADA and Corporate). If the Corporate network is disabled all server processes work fine and start up. If the corporate network is enabled all the server process do not startup well, exceptions ,not responding . Anyone experienced anything like this?

  • That is how it is configured 192 to SCADA ad 10. to historian.

    We stopped the connector at some point thinking it might be causing the issue but no

  • Gateway only defined on the CORP NIC, the SCADA NIC doesn't have one define

  • Are all SCADA server machines in your project having dual NICs and all of them configured like

    192.XXX.XXX.XXX,10.XXX.XXX.XXX in the Network Addresses field? Or you just use the machine names, not IPs. If dual NICs are just for Citect Connector, only the connector machine needs to have dual NICs and other SCADA shouldn't have any dual NICs. The problem is your webclient that is sitting on CORP network and required to directly connect SCADA servers for data requests. However, with dual NICs, the routing of IP traffic might not be desirable for SCADA applicaitons. As Eric pointed out, "Citect doesn't like to send a message on one network and receive a response from the other". A good solution is to get rid of dual NICs from server configuration as you are not using them for network redundancy and replace legacy webclient with PlantSCADA Access Anywhere.

  • Removed the webclient CORP IP from the project and all server processes and everything is good again. 
    Not sure why its kaput in this version, was okay in the previous. Who knows, thank you for your help

  • Removed the webclient CORP IP from the project and all server processes and everything is good again. 
    Not sure why its kaput in this version, was okay in the previous. Who knows, thank you for your help

  • One thing I noticed that I haven't seen commented on yet. You state that the SCADA traffic comes through on a NIC with a 192.XXX.XXX.XXX IP address, and the Historian traffic comes through on a NIC with a 10.XXX.XXX.XXX IP address. Well, 192.XXX.XXX.XXX is a Class C private IP address with a default Subnet Mask of 255.255.255.0, and 10.XXX.XXX.XXX is a Class A private IP address with a default Subnet Mask of 255.0.0.0. They are on different subnets, which means they comprise two, separate networks. If there is any information that is being passed into either the Historian or SCADA side from the opposite network, they will not see each other and the traffic will drop into the great IP void. Are you able to reassign IP addresses to put both NICS on the same network? Even if it's not a permanent change, if it works you'll have identified the nature of the problem.