cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
Announcements
469
Views
0
Helpful
2
Replies
Highlighted
Beginner

2811 with G.SHDSL WIC CPU running at 99%

I have a particular site that is causing me trouble, this site is connected in a back to back configuration using 2811 at CO and 2621XM at CPE.  The CO end is also the CO for 3 other sites so has a total of 4 wics installed (WIC-1SHDSL-v2), these other sites also have 2621XMs for the CPE.

The problem i am getting is when one site in particular transfers large files to/from client machines, the CPU on the 2811 jumps to 99%:

CPU utilization for five seconds: 99%/98%; one minute: 26%;

PID Runtime(ms)   Invoked      uSecs   5Sec   1Min   5Min TTY Process

  11    12881868  37249378        345  0.49%  0.50%  0.51%   0 ARP Input

  54     8548592  30375358        281  0.40%  0.45%  0.41%   0 XDSL BACKGROUND

  78     4947460  15276755        323  0.16%  0.15%  0.16%   0 IP Input

  77       31240 104868206          0  0.16%  0.15%  0.11%   0 IP ARP Retry Age

  72       36984 104868210          0  0.08%  0.09%  0.13%   0 ACCT Periodic Pr

As a result the latency for all 4 sites shoots through the roof and also dropping packets.  The same CPU utilisation is not shown at the CPE end, the CPU useage is still low.

Here is the config from the 2811 showing the problem:

controller DSL 0/3/0

mode atm

line-term co

line-mode 2-wire line-zero

dsl-mode shdsl symmetric annex B

line-rate auto

interface ATM0/3/0

  bandwidth 2196

no ip address

no ip route-cache cef

no ip route-cache

no atm ilmi-keepalive

!

interface ATM0/3/0.1 point-to-point

bandwidth 2196

ip address 172.25.200.241 255.255.255.252

no ip route-cache

no ip mroute-cache

cdp enable

pvc 0/32

  vbr-nrt 2188 2188 1

  no broadcast

  encapsulation aal5snap

  service-policy output Torbay_VOIP

The DSL is configured to auto negotiate and is running at 2312Kbps:

DSL 0/3/0 controller UP

Globespan xDSL controller chipset

Line Mode: Two Wire

DSL mode: Trained with SHDSL Annex B

Frame mode: Utopia

Configured Line rate: Auto

Line Re-activated 26 times after system bootup

LOSW Defect alarm: ACTIVE

CRC per second alarm: ACTIVE

Line termination: CO

FPGA Revision: 0xB3

       Current 15 min counters

               CRC : 0 LOSW Defect : 0 ES : 0  SES : 0 UAS : 0

       Previous 15 min counters

               CRC : 0 LOSW Defect : 0 ES : 0  SES : 0 UAS : 0

       Current 24 hr counters

               CRC : 11        LOSW Defect : 8 ES : 2  SES : 0 UAS : 91

       Previous 24 hr counters

               CRC : 0 LOSW Defect : 0 ES : 0  SES : 0 UAS : 0

Line-0 status

Chipset Version:  0

Firmware Version:  R4.2.1

Modem Status:  Data, Status 1

Last Fail Mode:  No Failure status:0x0

Line rate:  2312 Kbps

Framer Sync Status: In Sync

Rcv Clock Status: In the Range

Loop Attenuation:  12.5 dB

Transmit Power:  14.5 dB

Receiver Gain:  28.5420 dB

SNR Sampling:  36.3590 dB

Receive HEC Error Count:  0

Dying Gasp: Present

Any ideas how i can track down what is causing such high useage?

Regards

Jeff

Everyone's tags (5)
1 ACCEPTED SOLUTION

Accepted Solutions
Cisco Employee

2811 with G.SHDSL WIC CPU running at 99%

Hi,

You are seeing high CPU due to interrupts. Basically your CPU is being interrupted by the interface drivers and is asked to do the job of packet switching and this is being done frequently due to high rate of traffic.

Do you see this high CPU usage only with huge file transfers?

You are seeing high cpu due to interrupts.

http://www.cisco.com/en/US/products/hw/routers/ps359/products_tech_note09186a00801c2af0.shtml#cd

To start with can you turn on CEF on all the interfaces. I see CEF turned off on the WAN interface

interface ATM0/3/0.1 point-to-point

bandwidth 2196

ip address 172.25.200.241 255.255.255.252

no ip route-cache                >>>>>>>>

no ip mroute-cache            >>>>>>>>

http://www.cisco.com/en/US/products/sw/iosswrel/ps1828/products_tech_note09186a00801e1e46.shtml

Check if CPU reduces. If not, please provide the following info.

- CPU model

- 'show interface summ' when CPU is high

- 'sh run' if possible else 'show ip int'

Thanks.

2 REPLIES 2
Cisco Employee

2811 with G.SHDSL WIC CPU running at 99%

Hi,

You are seeing high CPU due to interrupts. Basically your CPU is being interrupted by the interface drivers and is asked to do the job of packet switching and this is being done frequently due to high rate of traffic.

Do you see this high CPU usage only with huge file transfers?

You are seeing high cpu due to interrupts.

http://www.cisco.com/en/US/products/hw/routers/ps359/products_tech_note09186a00801c2af0.shtml#cd

To start with can you turn on CEF on all the interfaces. I see CEF turned off on the WAN interface

interface ATM0/3/0.1 point-to-point

bandwidth 2196

ip address 172.25.200.241 255.255.255.252

no ip route-cache                >>>>>>>>

no ip mroute-cache            >>>>>>>>

http://www.cisco.com/en/US/products/sw/iosswrel/ps1828/products_tech_note09186a00801e1e46.shtml

Check if CPU reduces. If not, please provide the following info.

- CPU model

- 'show interface summ' when CPU is high

- 'sh run' if possible else 'show ip int'

Thanks.

Beginner

2811 with G.SHDSL WIC CPU running at 99%

Thanks for that, turning on CEF on the WAN interface dropped the CPU right back down again when transferring large files.

Regards

CreatePlease to create content
Content for Community-Ad
July's Community Spotlight Awards