08-01-2013 08:25 AM
I'd like to know what are the acceptable limits for the ratio between frames / frame error?
The fabos is 6.4.3d.
Thank you for your return.
08-01-2013 10:39 AM
Switching to the advanced view will produce a better split of the errors.
That said the statistics from the bottom screenie don't look good, however it's hard to tell what could be the cause.
Please clear the port statistics, wait an X amoun of time depending on the traffic and take some screenies, or post the portstats(64)show output.
08-05-2013 04:18 AM
I have reset all port and 72 hour later I have just one port who have error.
What do you think about this values :
3: 654.3m 260.2m 0 0 0 0 0 0 5.1k 0 0
It's an ESX server on this port.
08-05-2013 06:52 AM
just another question:
from the TX Buffer Zero, it looks like you have a long distance here.
Is there any long distance equipment e.g DWDM or so between your switches?
08-05-2013 06:55 AM
what switch type is that?
do the following
statsclear on the switch
then reissue the portstatsshow command like every 1-2 seconds and see if the counter time_txcrd_z increases and if yes, with what rate does it do that?
issue portcfgshow (portnumber) just want to see the fillword setting and ifts set correct. check with the hba vendor what fillword setting has to be applied to the port
08-05-2013 08:39 AM
Schlegel, the port number 3 is connected directly to a ESX server, the optical cable is between the optical sw and the server is 5 ft ( 1,5 meter).
Sorry but what mean : "e.g DWDM" ?
Doges, The type of switch is Brocade 300, IBM vendor.
After clearing stats, tim_txcrd_z increase around 10 000 every sec.
portcfgshow 3 retun this value :
Fill Word: 3(A-A then SW I-A)
On an other switch with the same configuration connected, I have no error and the port have the same configuration. Althought the tim_txcrd_z stay at 0.
You say I must call the vendor (DELL ? emulex fc card Or VMWare ?) to get the fillword setting ?
Thank for your help
If it can help all the error are classified like this :