TFTP Traffic issue on Cisco 3750

Unanswered Question
Jul 13th, 2010


I have a very bizarre issue. TFTP transfers are very slow when I have my server hooked up to my Cisco 3750. Transfers were working fine a few weeks ago. However, if I hook my server up to a netgear 5 port unmanaged switch, and then the netgear to the 3750 tftp transfers work fine.

TFTP Dead Slow: Server -> Cisco 3750 <- Client [Both server and client are connected to the 3750]

TFTP Normal: Server -> Netgear -> Cisco 3750 <- Client [Client is hooked up to 3750, where as the server is connected to the netgear, and the netgear is connected to the 3750]

All other services on the server (ssh, nfs, etc...) work as expected and I don't have any network holds ups.

Also, I introduced another tftp server on the network and that is connected directly into the switch stack. I don't have any problems with using that tftp server.

I will post my switch configuration a little later, but I just have trunk links configured and 3 vlans setup with vtp.

I am very confused as to why a netgear 5 port unmanaged switch is fixing my problems.

Any help would be highly appreciated.



I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Amit Singh Tue, 07/13/2010 - 14:16


Didi you check the interface stats  on the 3750 when the server was connected to it. It might be dute to speed/ duplex mismatch on the interface when the server was connected. Sometime hardware problems create these kind of issues.It could also be due to bad patch cable connecting the server and the switch. Did you give a check to all that?


-amit singh

mgyurgyik Tue, 07/13/2010 - 14:33

Thank you for the quick reply.

Both the server and the switch port autonegotiated full duplex at 1000Mb/s. I also tried manually setting full-duplex + 1000Mbps on both the switch and the server with no luck. I'm pretty sure the cable is ok, since I am using the same cable to connect the server to the netgear switch. Also, other services such scp, nfs, etc... work fine at normal speeds.

I did try another patch cable for kicks without any improvments.

mgyurgyik Tue, 07/13/2010 - 15:37

MDF_CORE#show interfaces gigabitEthernet 2/0/19
GigabitEthernet2/0/19 is up, line protocol is up (connected)
  Hardware is Gigabit Ethernet, address is 0026.5233.c913 (bia 0026.5233.c913)
  MTU 1500 bytes, BW 1000000 Kbit, DLY 10 usec,
     reliability 246/255, txload 1/255, rxload 1/255
  Encapsulation ARPA, loopback not set
  Keepalive set (10 sec)
  Full-duplex, 1000Mb/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 00:00:42, output hang never
  Last clearing of "show interface" counters 00:01:27
  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 0 bits/sec, 0 packets/sec
  5 minute output rate 0 bits/sec, 0 packets/sec
     828 packets input, 426247 bytes, 0 no buffer
     Received 11 broadcasts (6 multicasts)
     0 runts, 0 giants, 0 throttles
     82 input errors, 82 CRC, 0 frame, 0 overrun, 0 ignored
     0 watchdog, 6 multicast, 0 pause input
     0 input packets with dribble condition detected
     1006 packets output, 67923 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 PAUSE output
     0 output buffer failures, 0 output buffers swapped out

I take it these errors might be a contributing factor to my problem?

Nagaraja Thanthry Tue, 07/13/2010 - 15:44


You are seeing CRC errors on the interface. This means that there is some sort of hardware incompatibility issue or the devices are not negotiating speed/duplex/other link parameters properly. As a first step, I would change the cable to a known good cable (If there are any patch panels along the path, bypass them). The next step would be to check the server side to see if there are any parameters that are hard coded (in addition to speed/duplex). Sometimes there are driver issues with certain hardware that result in these kind of issues. Please check the latest driver availability for the Ethernet card.

Hope this helps.



mgyurgyik Tue, 07/13/2010 - 17:01

Alright I'm pretty sure I have a bad nic card.

I've tried about 3-4 different cables all with the errors and the orginal cable I was using, I knew work. That cable hasn't been moved since I setup the server. In addition, if I set the speed on the switchport to 100Mbps all is good. As soon as I bump it up to 1000Mbps we have problems.

Edit: I am using linux so of course there aren't newer nic modules so to speak and I checked and I haven't had a kernel upgrade since I've installed debian on this box.

Thanks to all who have helped!


This Discussion

Related Content