Catalyst 2960s Problems

[ciscomagu]

Hi,

I wounder if there is any solution about high CPU, drops and packet loss on Cisco Catalyst 2960s yet.

We are using  4 four 2960s in a stack environment and have some problem. The CPU is high and we have port-asic drop and packet loss. When I search on Internet, there are many with same problem. This should be a big issue for Cisco to solve!

Supervisor TxQueue Drop Statistics

Queue  0: 29053

Queue  1: 0

Queue  2: 0

Queue  3: 103431

Queue  4: 0

Queue  5: 0

Queue  6: 0

Queue  7: 0

Queue  8: 579416

Queue  9: 0

Queue 10: 0

Queue 11: 0

Queue 12: 0

Queue 13: 0

Queue 14: 10478748

Queue 15: 0

8888889989899888889888888888889888888988888898888998899889888898898989

7955594760921878562798784798557757576469487928578506999769775537728153

100        *                      *                  *   **  *

90 ************************ *************** *****************************

80 #*#####*#############################################**##*############

70 #*###################################################**###############

60 #*###################################################**###############

50 ######################################################################

40 ######################################################################

30 ######################################################################

20 ######################################################################

10 ######################################################################

0....5....1....1....2....2....3....3....4....4....5....5....6....6....7.

0    5    0    5    0    5    0    5    0    5    0    5    0

CPU% per hour (last 72 hours)

• = maximum CPU%   # = average                                    CPU%

Many thanks for any comments

Best Regards

/Magnus

[i.harvey]

Now running ver 12.2(55)SE - definitely less packet drops.

Many thanks

(eagerly waiting for (58) now)

regards

Ian

[i.harvey]

The problem of packet loss manifested itself by affecting VOIP calls.

With packet loss being intermittent, VOIP calls would suddenly have one way traffic for a short period and then would return to normal. Now using ver 12.2(55) as opposed to 12.2(53), there have been no reports of this happening today.

The problems did not occur immediately with ver 12.2(53), only after ~ 3 weeks. the switches have been rebooted now due to the IOS upgrade, so hopefully it is not the reboot that has 'fixed' it for now

[dixho]

If the problem occurs again,  I suggest that you collect the following output:

show platform port-asic stats drop

show interface count

show spanning tree vlan

Most problems like that is resolve by enable QoS and priority output queue. However, we need to make sure that the voice traffic is marked with DSCP 46 or CoS 5 by the switch or the VOIP device.

[Leo Laohoo]

12.2(58)SE is now out (08 April 2011).