ifconfig overruns on a 3com

i've noticed on one of my interfaces a lot of overruns :
eth1 Link encap:Ethernet HWaddr 00:04:75:E7:C4:24
inet addr:10.0.0.1 Bcast:10.0.0.255 Mask:255.255.255.0
inet6 addr: fe80::204:75ff:fee7:c424/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:10802596 errors:0 dropped:0 overruns:104450 frame:0
TX packets:11807483 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:100
RX bytes:2506778114 (2390.6 Mb) TX bytes:287683394 (274.3 Mb)
Interrupt:16 Base address:0xd000

it's a 3com
this iface is connected to a switch,
02:04.0 Ethernet controller: 3Com Corporation 3c905C-TX/TX-M [Tornado] (rev
78)
mii output
# mii-diag eth1
Basic registers of MII PHY #24: 3000 782d 0041 6800 05e1 45e1 0007 2801.
The autonegotiated capability is 01e0.
The autonegotiated media type is 100baseTx-FD.
Basic mode control register 0x3000: Auto-negotiation enabled.
You have link beat, and everything is working OK.
Your link partner advertised 45e1: Flow-control 100baseTx-FD 100baseTx
10baseT-FD 10baseT, w/ 802.3X flow control.
End of basic transceiver information.

What can cause these problems. This interface is used in a thinclient
environment, and i noticed once in a while that the thinclients didn't
respond anymore.

Jef Peeraer <jef.peeraer@deadspam.com> wrote:
> i've noticed on one of my interfaces a lot of overruns :
> eth1 Link encap:Ethernet HWaddr 00:04:75:E7:C4:24
> inet addr:10.0.0.1 Bcast:10.0.0.255 Mask:255.255.255.0
> inet6 addr: fe80::204:75ff:fee7:c424/64 Scope:Link
> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
> RX packets:10802596 errors:0 dropped:0 overruns:104450 frame:0
> TX packets:11807483 errors:0 dropped:0 overruns:0 carrier:0

> What can cause these problems. This interface is used in a thinclient
> environment, and i noticed once in a while that the thinclients didn't
> respond anymore.

Seems that you have a station that is "Jabbering" (ie. It won't shut up
when its supposed to, and is sending packets that are too large.)

You said that you are connected via a hub. Using a switch will contain
the problem (if its store-and-forward anyway).

Managed _hub_ will sometimes have a feature called Overrun protection,
or Jabber Latch. It shuts down a port when it decides that a port is
receiving too many errorred frames.

See Ethernet: The Definitive Guide for more information.

To troubleshoot this scenario, I suggest you run "watch ifconfig eth0"
on the machine, then take out each client on the network for however
long it takes for you to notice that the count is not increasing. (It it
just slows down, there may be multiple jabberers.

--
Cameron Kerr
cameron.kerr@paradise.net.nz : http://nzgeeks.org/cameron/
Empowered by Perl!

Cameron Kerr wrote:

> Jef Peeraer <jef.peeraer@deadspam.com> wrote:
>> i've noticed on one of my interfaces a lot of overruns :
>> eth1 Link encap:Ethernet HWaddr 00:04:75:E7:C4:24
>> inet addr:10.0.0.1 Bcast:10.0.0.255 Mask:255.255.255.0
>> inet6 addr: fe80::204:75ff:fee7:c424/64 Scope:Link
>> UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
>> RX packets:10802596 errors:0 dropped:0 overruns:104450 frame:0
>> TX packets:11807483 errors:0 dropped:0 overruns:0 carrier:0
>
>> What can cause these problems. This interface is used in a thinclient
>> environment, and i noticed once in a while that the thinclients didn't
>> respond anymore.
>
> Seems that you have a station that is "Jabbering" (ie. It won't shut up
> when its supposed to, and is sending packets that are too large.)
the only pc connected is a thinclient ( epia ve 5000 ) with a via rhine
chipset. Maybe i have to seek in that direction.
>
> You said that you are connected via a hub. Using a switch will contain
> the problem (if its store-and-forward anyway).
sorry, i didn't express myself correctly. i am using a switch ( nway
24-port ).
>
> Managed _hub_ will sometimes have a feature called Overrun protection,
> or Jabber Latch. It shuts down a port when it decides that a port is
> receiving too many errorred frames.
>
> See Ethernet: The Definitive Guide for more information.
i'll have this book, and i should have checked it . shame on me !
>
> To troubleshoot this scenario, I suggest you run "watch ifconfig eth0"
> on the machine, then take out each client on the network for however
> long it takes for you to notice that the count is not increasing. (It it
> just slows down, there may be multiple jabberers.

thanks for the info anyway!
>

Jef Peeraer <jef.peeraer@deadspam.com> wrote:
> Cameron Kerr wrote:

>> You said that you are connected via a hub. Using a switch will contain
>> the problem (if its store-and-forward anyway).
> sorry, i didn't express myself correctly. i am using a switch ( nway
> 24-port ).

Then perhaps you should see if you can substitute the switch for
another.

--
Cameron Kerr
cameron.kerr@paradise.net.nz : http://nzgeeks.org/cameron/
Empowered by Perl!