Issues Between Cisco Network & NVT Phybridge POLRE 8 Port Extender

[bmiller23]

Hello,

I currently have a network consisting of a Cisco 4510r-e switch and a number of Cisco 3750G POE Switches ties together via fiber.

IOS versions:
4510R-E - 15.0(2)SG7
3750G - 12.2(35)SE5

For stand alone testing:
3750G - IOS 15.0(2)SE2

NVT POLRE 8 Port Extender

What's happening is when I connect the POLRE to my 4510 switch on any port I experience tremendous packet loss when attempting to communicate with any device plugged into the POLRE. I experience the same loss when the POLRE is connected to any port on my 3750G switches. With the same port configuration I have tested the POLRE on a stand alone 3750 switch and everything works fine. It also works fine if I connect the POLRE to an unmanaged switch. However, as soon as I plug the stand alone 3570 with the POLRE connected into any of my switch open ports, I drop packets again. This is not a IP issue and I have an extremely simply network configuration running a single VLAN on a /28 IP routed network (255.255.240.0). I have tried turning the inline power off and cdp in the port configuration to no avail. NVT is at a loss and I don't have a current Cisco support contract.

Any suggestions would be greatly appreciated.

 

[iggsterman]

What's the configuration of interfaces on 4510 and 3750G that you tested with?

 

[bmiller23]

4510 -
Currently Testing:
switchport mode access
power inline never
speed 100
duplex full
no cdp enable

Original Config:
switchport mode access

3750G:
switchport mode access

 

[bmiller23]

I have now run wireshark while the POLRE is connected to my 4510 to see if I could see any issues in transmission / receipt from that IP. Trying to open a web page on the IP I am seeing many tcp retransmissions, spurious retransmissions, and tcp dup acks. Unfortunately I am a complete newbie on wireshark so I don't know to to troubleshoot this much further. If I don't change a thing and connect it to my stand alone 3750 I get no such errors.

If I eliminate this POLRE and plug my device directly to the 4510 I have no errors, so there definitely is something going on between my 4510 and the POLRE.

 

[iggsterman]

On the 4500:

no speed 100
no duplex full

You are connecting a hard set 100/full to an auto-neg 1gig - BAD idea.
As for CDP I doubt it would do any harm.

 

[iggsterman]

On another note, have you looked at the output of
sh interface _your port intf_
on the 4510? That could provide a lot of clues.

 

[bmiller23]

I agree on the speed setup and the duplex settings. I simply tried them as this POLRE wants 100Mbps and full duplex. I have removed those settings. here is the sh interface and some ping stats:

CCGRC#show interfaces gi1/47
GigabitEthernet1/47 is up, line protocol is up (connected)
Hardware is Gigabit Ethernet Port, address is 001f.6cdc.d47e (bia 001f.6cdc.d47e)
MTU 1500 bytes, BW 100000 Kbit, DLY 100 usec,
reliability 255/255, txload 113/255, rxload 1/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, 100Mb/s, link type is auto, media type is 10/100/1000-TX
input flow-control is off, output flow-control is off
ARP type: ARPA, ARP Timeout 04:00:00
Last input 17:51:43, output never, output hang never
Last clearing of "show interface" counters 00:43:53
Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 5511
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 0 bits/sec, 0 packets/sec
5 minute output rate 44511000 bits/sec, 4295 packets/sec
1121 packets input, 847977 bytes, 0 no buffer
Received 30 broadcasts (24 multicasts)
0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 input packets with dribble condition detected
11240427 packets output, 14538438025 bytes, 0 underruns
0 output errors, 0 collisions, 0 interface resets
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier
0 output buffer failures, 0 output buffers swapped out

Ping statistics for 192.168.96.102:
Packets: Sent = 127, Received = 48, Lost = 79 (62% loss),
Approximate round trip times in milli-seconds:
Minimum = 4ms, Maximum = 27ms, Average = 8ms

The only thing I see there is some output drops. There is no congestion on this interface and all other interfaces on this card are not experiencing issues. I have replaced the cables and even had NVT RMA me a new POLRE and had the same results.

 

[iggsterman]

Maybe a silly question, you did plug 4510 into the uplink port on POLRE and not one of those 8 ports? It shows it as negotiated 100/full and it should be 1000 full. Also, you said POLRE wants 100 full? How so? I thought the uplink ports are auto-neg 1000/100/10 and the extender ports are 100/10? There is some traffic on your port however: your interface shows about 50% load on TX. What else is going on there?

 

[bmiller23]

According to what I have read the uplink ports on the POLRE are 100 full duplex. The extender ports are 10 full duplex. From the spec sheet:

Interface: Ethernet Uplink (Trunk IP)
2 RJ45 ports: 10/100 Base-T auto-sensing, independent speed selection, Ethernet IEEE 802.3, CAT5 copper cable
Interface: Downlink (PoE and IP to Adapter)

8 x RJ11 Jacks
Maximum Distance: 1200ft (365 m) CAT3 UTP cable,
24 AWG
Speed: 10 Mbps (full duplex)

I have the uplink plugged into interface gi1/47. I have one IP camera plugged into the first extender port. There is nothing else running on this interface. To test the camera, I moved it to interface gi1/45. Here's the stats and it works perfectly:

CCGRC#sho interfaces gi1/45
GigabitEthernet1/45 is up, line protocol is up (connected)
Hardware is Gigabit Ethernet Port, address is 001f.6cdc.d47c (bia 001f.6cdc.d47c)
MTU 1500 bytes, BW 100000 Kbit, DLY 100 usec,
reliability 255/255, txload 104/255, rxload 1/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, 100Mb/s, link type is auto, media type is 10/100/1000-TX
input flow-control is on, output flow-control is on
ARP type: ARPA, ARP Timeout 04:00:00
Last input 18:39:37, output never, output hang never
Last clearing of "show interface" counters 00:12:17
Input queue: 0/2000/0/0 (size/max/drops/flushes); Total output drops: 0
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 0 bits/sec, 0 packets/sec
5 minute output rate 40903000 bits/sec, 3981 packets/sec
12200 packets input, 17925251 bytes, 0 no buffer
Received 9 broadcasts (8 multicasts)
0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 input packets with dribble condition detected
3166673 packets output, 4092446857 bytes, 0 underruns
0 output errors, 0 collisions, 0 interface resets
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier
0 output buffer failures, 0 output buffers swapped out

 

[bmiller23]

So for a comparison I plugged the POLRE into a stand alone switch (nothing else running and not connected to any device). The camera worked fine and here are the stats:

GigabitEthernet1/0/3 is up, line protocol is up (connected)
Hardware is Gigabit Ethernet, address is 0023.5d74.0683 (bia 0023.5d74.0683)
MTU 1500 bytes, BW 100000 Kbit, DLY 100 usec,
reliability 255/255, txload 1/255, rxload 15/255
Encapsulation ARPA, loopback not set
Keepalive set (10 sec)
Full-duplex, 100Mb/s, media type is 10/100/1000BaseTX
input flow-control is off, output flow-control is unsupported
ARP type: ARPA, ARP Timeout 04:00:00
Last input never, output 00:00:00, output hang never
Last clearing of "show interface" counters never
Input queue: 0/75/0/0 (size/max/drops/flushes); Total output drops: 0
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 6033000 bits/sec, 564 packets/sec
5 minute output rate 104000 bits/sec, 212 packets/sec
2229643 packets input, 2930990006 bytes, 0 no buffer
Received 44 broadcasts (0 multicasts)
0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 watchdog, 26 multicast, 0 pause input
0 input packets with dribble condition detected
863497 packets output, 55326628 bytes, 0 underruns
0 output errors, 0 collisions, 1 interface resets
0 babbles, 0 late collision, 0 deferred
0 lost carrier, 0 no carrier, 0 PAUSE output
0 output buffer failures, 0 output buffers swapped out

So something on my network is showing traffic through on my interface as the bandwidth the camera and POLRE is actually using is 5.75Mbit, not 39Mbit. I think I'm misunderstanding something here.

 

[bmiller23]

I figured it out. In looking at the number of drops I was pretty sure there was another culprit. What I tried was issuing the following command to the POLRE port:

switchport block multicast

Once I did that the camera and port instantly started working. I am now going to have to find the device that is flooding my network with unnecessary multicast packets. If anyone has any ideas on that I would be greatly appreciative.

 

[ADB100]

Any MS NLB servers running.....?

 

[bmiller23]

No. MS Server 2012 and a bunch of Windows Embedded Standard clients.

 

[VinceWhirlwind]

You have a look for the multicast IP addresses that your switches can see and then google them - they are unique to their vendor.