Emerson Exchange 365

• Not Answered

ACN COMM EVENTS for random stations and random time.

Hi, 

While browsing through events we can see multiple stations switching to Secondary ACN and again Switching back to primary ACN within fraction of seconds. This happens at random intervals on random stations.We are not using any smart switch. What might the issue. Nobody has reported that graphics going bad till date but this issue has been going on for a while.

• DeltaV #DCS

Neeraj S

• 15 Dec 2020 8:03 AM
• Cancel

2 Replies

• Andre Dicaire

  

  • 19 Jan 2021 5:43 PM

  An ACN switchover is triggered when a node does not receive a response within the allotted timeout and retry period. Since the DeltaV network prefers to run on the Primary, a switch to the secondary will be followed by a switch back to the Primary once the Primary network is confirmed good.
  
  The DeltaV Smart switches play no role in this. The logic is strictly managed by the nodes. Smart Switches offer some cybersecurity features that protect against broadcast and multicast storms, as well as locking down ports to defined MAC address of connected devices and reporting issues into the Event Chronicle and Hardware Alerts. If yo use CHARM IO Cards, these should be connected to their controllers via Smart Switches. The FW/IPD device can also provide network segregation and protection from cyber threats and broadcast/Multicast storms.
  
  DeltaV uses unicast messaging to send normal process traffic, which means individual packets are sent to each destination that has requested data, such as consoles and peer to peer connections. Each packet sent triggers a confirmation packet that it was received. If the confirmation is not received within the Timeout *Retry period, that connection to that specific node is considered lost and the node switches to the Secondary to continue communications. All other connections would remain on the Primary unless they too experience the time out. Normally all communications occur on the primary, but based on network connectivity, the controller can maintain any mix of connections via primary and/or secondary.
  
  If there is a network problem, like a bad cable, Switch port, or network storm, and this intermittently causes a loss of connection, an ACN switchover will occur. Subsequent diagnostic packets on the primary subsequently respond and communication switches back to the primary.
  
  Other reasons could be that a console is rebooted or DeltaV services are temporarily stopped. Or possibly the communication layer is unable to respond to messages long enough to force a time out at the source.
  
  The ACN messages should be reviewed as a whole to isolate the common node or network segment/component, giving an idea of where to look. Some say that having some ACN errors are normal, even on a healthy system. I'd agree, only in that even a Normal system can experience passing disruptions such as a console reboot, controller switchovers etc. The ACN switchover happens due to time out on packet confirmation.
  
  If you permanently disconnect a node from the system, you do need to properly remove it by downloading Setup data. The Node Table tells all nodes about the node names and IP addresses. Once a controller connects to a node it adds this to its Connection List. you can view this via diagnostics, right click on Communications and select View Device Connections. Once in the list, it remains there. This triggers ACN switchover messages if the node is disconnected and was in the connection list. Interestingly, if you run Diagnostics on a workstation, it will connect to all nodes and appear in this list. Otherwise, nodes only appear if they explicitly requested data from the controller, like pulling up a display that calls for data from a module in that controller. Before removing the node, decommission and/or delete it from Explore and download the changed setup data to all nodes. Then you can disconnect the node in question. The node will still appear in the connection list of controllers and/or IO nodes but will be closed. To remove the node from this list, you can perform a double switchover if the node is redundant. The next time you update firmware, or if there is a power failure, the node will be removed.
  
  This probably more than you wanted to know about ACN switchovers. I would at least review them to see if there is any common bad actor or network segment involved. If you can correlate to downloads or reboots and such, they are "normal".

  Andre Dicaire

  • Cancel
• Neeraj S

  

  • 27 Jan 2021 5:47 AM

  In reply to Andre Dicaire:

  Thanks Andre for the information.

  • Cancel