RuggedCom RS400 用户手册

下载
页码 275
Spanning Tree 
 
 
ROS™  v3.5 
166 
RS400 
5.6 Troubleshooting 
Problem One 
When I connect a new port the network locks up. The port status LEDs are flashing 
madly. 
Occasionally, the network seems to experience a lot of flooding. All the ports seem to 
experience significant traffic. The problem lasts a few seconds and then goes away.  
One of my switches displays a strange behavior where the root port hops back and forth 
between two switch ports and never settles down.  
Is it possible that one of the switches in the network or one of the ports on a switch in the 
network has STP disabled and accidentally connects to another switch?  If this has occurred 
then a traffic loop has been formed. 
If the problem appears to be transient in nature, it is possible that ports that are part of the 
spanning tree have been configured as edge ports. After the link layers have come up on edge 
ports, STP will directly transition them (perhaps improperly) to the forwarding state. If an RSTP 
configuration message is then received the port will be returned to blocking. A traffic loop may 
be formed for the length of time the port was in forwarding. 
If one of the switches appears to flip the root from one port to another the problem may be one 
of traffic prioritization (See problem five).  
Another possible cause of intermittent operation is that of an auto-negotiation mismatch. If one 
end of the link is fixed to full duplex and the peer auto-negotiates, the auto-negotiating end will 
fall back to half-duplex operation. At lower traffic the volumes the link may display few if any 
errors. As the traffic volume rises, the fixed negotiation side will begin to experience dropped 
packets while the auto-negotiating side will experience collisions. Ultimately, as traffic loads 
approach 100% the link will become entirely unusable. At this point RSTP will not be able to 
transmit configuration messages over the link and the spanning tree topology will break down. If 
an alternate trunk exists RSTP will activate it in the place of the congested port. Since activation 
of the alternate port often relieves the congested port of its traffic, the congested port will once 
again become reliable. RSTP will promptly enter it back into service, beginning the cycle once 
again. The root port will flip back and forth between two ports on the switch.  
Problem Two 
My PC/IED/Device is connected to your switch. After I reset the switch, it takes a long 
time before it comes up. 
Is it possible that the RSTP edge setting for this port is set to false? If edge is set false the 
bridge will make the port go through two forward delay times before the port can send or receive 
frames. If edge is set to true the bridge will transition the port directly to forwarding upon link up. 
Another possible explanation is that some links in the network run half-duplex. RSTP uses a 
peer-peer protocol called Proposal-Agreement to ensure transitioning in the event of a link 
failure. This protocol requires full duplex operation. When RSTP detects a non-full duplex port it 
cannot rely on Proposal-Agreement protocol and must make the port transition the slow (i.e. 
STP) way. If possible, configure the port for full-duplex operation otherwise configure the port’s 
point-to-point setting to true. Either one will allow the Proposal-Agreement protocol to be used.