6509 - 10Gb DFC3C - 48 10/100/100 6700-CFC

Answered Question
Jul 29th, 2010
User Badges:

I inserted online (OIR) a 10GB card into one 6509 with SUP720 & PFC3B.

during card insertion, I got the special msg :

"%OIR-SP-6-DOWNGRADE_EARL: Module 1 DFC installed is not identical to system PFC and will perform at current system operating mode"

this message is normal and indicates the PFC3B and the DFC3C are nbot of the same kind, and the DFC3C will work on downgraded mode.


My problem is the following :


Since the 10Gb was inserted, all my 48-port 10/100/1000 card (6700-CFC cards) experience a trafic overflow problem in TxLoad, specially for

ports configured at 100 Mb/s. And the servers connected at 100 Mb/s to these cards are suffering of low performances.


Can anybody have any idea ?  I didn't try to reboot the switch up-to-now, because I can do it only during out-of-normal hours.


By the way I have an other 6509 with exactly the same hardware configuration, and it is performing well.


Pls find in the "shwo6509.doc" file attached, the following displays :


a - "show modules"

b - the messages I got during the insertion of the 10Gb card.

c - "sh int status | inc 9/"

d - "sh int g 9/29


the last display (d-) shows a TxLoad of 249/255 and this is the same on all 100Mb/s port of this card and for all other 48-port 10/100/1000 cards in this chassis.

Attachment: 
Correct Answer by gephelps about 6 years 11 months ago

I would use SPAN on one of the ports and see what the traffic is. The txload is high because the port is sending out a ton of traffic:


/* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:"Times New Roman"; mso-ascii-font-family:Cambria; mso-ascii-theme-font:minor-latin; mso-fareast-font-family:"Times New Roman"; mso-fareast-theme-font:minor-fareast; mso-hansi-font-family:Cambria; mso-hansi-theme-font:minor-latin; mso-bidi-font-family:"Times New Roman"; mso-bidi-theme-font:minor-bidi;}

S6509-PYR-SS-02#sh int g 9/29
GigabitEthernet9/29 is up, line protocol is up (connected)

Last clearing of "show interface" counters 1y0w

5 minute input rate 0 bits/sec, 0 packets/sec
5 minute output rate 97986000 bits/sec, 12451 packets/sec

500484896 packets input, 147074385347 bytes, 0 no buffer

8790471090 packets output, 8626481548805 bytes, 0 underruns


The counters have not been cleared for over a year so I ignored a lot of the error counters because we do not know when they were incremented. For the last five minutes though there have been zero packets input yet we are sending out 12,451 packets per second. The number would probably be larger if this port was configured for gigabit.




  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 5 (1 ratings)
Loading.
Correct Answer
gephelps Thu, 07/29/2010 - 14:52
User Badges:
  • Cisco Employee,

I would use SPAN on one of the ports and see what the traffic is. The txload is high because the port is sending out a ton of traffic:


/* Style Definitions */ table.MsoNormalTable {mso-style-name:"Table Normal"; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-parent:""; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:12.0pt; font-family:"Times New Roman"; mso-ascii-font-family:Cambria; mso-ascii-theme-font:minor-latin; mso-fareast-font-family:"Times New Roman"; mso-fareast-theme-font:minor-fareast; mso-hansi-font-family:Cambria; mso-hansi-theme-font:minor-latin; mso-bidi-font-family:"Times New Roman"; mso-bidi-theme-font:minor-bidi;}

S6509-PYR-SS-02#sh int g 9/29
GigabitEthernet9/29 is up, line protocol is up (connected)

Last clearing of "show interface" counters 1y0w

5 minute input rate 0 bits/sec, 0 packets/sec
5 minute output rate 97986000 bits/sec, 12451 packets/sec

500484896 packets input, 147074385347 bytes, 0 no buffer

8790471090 packets output, 8626481548805 bytes, 0 underruns


The counters have not been cleared for over a year so I ignored a lot of the error counters because we do not know when they were incremented. For the last five minutes though there have been zero packets input yet we are sending out 12,451 packets per second. The number would probably be larger if this port was configured for gigabit.




MARC MONTANARI Mon, 08/02/2010 - 02:42
User Badges:

For a performance problem, I did a SPAN and captured traffic with WireShark,

and I figured out my 6509 was flooding the trafic destined to server 10.1.100.1 (this is a big NetApp server)


This server is on port Ten1/2 of C-6509_02.  (I have 2 x C-6509 with the same hardware configuration).


I checked the mac-address of the server in the "show mac-address dynamic" display, and I discovered this mac-address is not learned by the switch.


Consequently, in order to prevent this flooding, I configured a mac-address static on port Ten1/2, with the mac-address of this server.

The flooding of the traffic to 10.1.100.1 stopped immediatly.


My 2 x C-6509 have exactly the same IOS release and the same hadrware configuration.


On port Ten1/2 of C-6509_01, I have a server PYRBU06 and its mac-address is well registered dynamically.

On port Ten 1/2 of C-6509_02, I have a server PYRBU05, of which the mac-address cannot be learned by the switch.


Anybody has an idea of why my C-6509_02 doesn't learn the mac-address of server PYRBU05 ?


Thank you for any help

Marc


I join the configurations of both C-6509

Actions

This Discussion