Normal Amount of Discards

readymixed1 · ‎11-12-2008

What would you guys consider to be around a normal amount of Transmit Discards on a switch?

I'm asking this because we use solarwinds and under High Errors & Discards it's showing pretty much all my HP Thin Clients, Computers and some printers with discards over 500,000,000!!!

I would think that is not normal at all!!

But we have a stack of 7 switches and almost every port on all the switches are like that.

Any ideas?

ropethic · ‎11-12-2008

Check to see if jumbo frames is enabled. Sounds like if jumbo frames are enabled on devices they are being dropped on the switch ports.

show system MTU

500M discards is not normal

What version are you running? Platform?

readymixed1 · ‎11-12-2008

Show System MTU:

System MTU size is 1500 bytes

System Jumbo MTU size is 1500 bytes

Routing MTU size is 1500 bytes

Cataylst 3750 (stack of 7, 3 which are 10/100 and 4 which are 10/100/1000)

Cisco IOS Software, C3750 Software (C3750-IPBASE-M), Version 12.2(44)SE2

johnlloyd_13 · ‎11-12-2008

that number is insane! test your cables and replace to test it. you might also want to check and correct duplex and speed settings.

readymixed1 · ‎11-12-2008

I took 3 ports and check them. They are at Full Duplex Speed 100

So i set speed and duplex to auto and i'll see what happens.

Whats strange is it is all the ports on all the switches pretty much.

The network person before me set every single port to the following:

storm-control broadcast level 20

spanning-tree portfast

spanning-tree bpduguard enable

Could this be somewhat of a problem? Because we have through the building 3com 8 port switches that have multiple devices plugged into them.

Although I still need to figure out what is causing all the problems.

I'm sure some of the cables are old and we are in the process of switching them out but I just find it strange that almost all the ports report a high transmit discard (this is through Solarwinds).

I am in the process of setting up IDFs (1 for up stairs, 1 for the front of the building and 1 for the back of the building). Then running the IDfs back to our server room and into our 3750G's.

But back to the high discard packets and also the spanning-tree portfast on all the ports.

readymixed1 · ‎11-12-2008

Another note,

Would you disable Flow Control on your ports?

johnlloyd_13 · ‎11-12-2008

is the report you received from solarwinds set at 24 hours or 7 days interval? how about polling of your devices? you might also want to check your NMS settings.