Ethernet IP troubleshooting

[The Plc Kid]

Ken I was basing firmware / multicast on the fact that he is using a L61 so he could be using any firmware 15-20 IIRC.

May be worth flasing it up to 19 or so to Have Unicast.

 

[The Plc Kid]

Whether the network has a multicast flood or not, the ENBT is handling it with ease - 35% utilisation is what it says...

35% cpu utilisation does not mean you don't or can't have a multicast traffic issue. On computer networks you can have a multicast flooding issue and your desktops and server nic's cpu's arenot at a 100%.

EIP hardware works the same way. The ENBT is a NIC for the L61.

 

[paulB]

ENBT module firmware revision is 6.4 from Module Info Tab on properties. However on GENERAL tab it is 4.1. I'll try to match it and reload program.

 

[paulB]

Drives are displaying F075 fault.
Processor is rev. 17 by customer request. I am not sure how to set everything to unicast versus multicast.

 

[The Plc Kid]

I don't think it is a option in 17. I think it became a option in 18 and the default in 19 up IIRC.

 

[The Plc Kid]

May want to check the firmware in your 20-COMM cards and compare to the release notes for the given firmware and above.

 

[The Plc Kid]

Do you have access to a managed switch? Might be worth a try.

 

[paulB]

No we do not have managed switch available.

 

[paulB]

I am looking at ENBT web page on Ethernet statistics. I am not sure how to post a snap shot.
However all media counters are 0 exept FCS Errors. It equals 7718 and increments by 6 once in awhile.

 

[paulB]

Fault on powerflex 40 is f071

 

[Kick-It-Up]

As was already mentioned check your link speed and duplex settings, set every device to auto negotiate. Also check your firmware revisions for updates due to networking issues, your getting framing errors so those are you two most likely culprits. Also as was suggested flashing up to a firmware that supports unicast would be advisable particularily with an unmanaged switch.

 

[daba]

FCS errors can be caused by fall-back to half-duplex...

read the whole thread, it's worth it

 

[OkiePC]

I am in agreement with the Kid on this one.

Just because the enbt can handle it, doesn't mean the switching fabric of the unmanaged switch can handle broadcasting it. There are probably still adjustments that can be made at the switch level, we have RS20 unmanaged and managed switches of various firmwares. I highly recommend unicast since you can't insert a managed switch.

Be wary of certain spanning tree protocols as well.

FCS errors, in my limited experience, do point to the probability that some device on the network is doing auto-negotiate messaging. This needs to be forced, in my current opinion. it's not necessary at the control hardware level, and can not help you as much as it can jump up and bite you.

 

[Ken Roach]

A round of drinks on me if the OP comes back saying the switch in unmanaged (and therefore always has Autonegotiate enabled) but that "somebody at Rockwell" told him to hard-set the device ports to 100/Full.

FCS errors point to a duplex mismatch, bad Ethernet cable or termination, or a noise problem.

 

[paulB]

Switch is unmanaged and all device ports are set to 100/full. I'll switch it to autonegotiate tomorrow and see if it helps.