[c-nsp] High CPU caused by interupt on 7600 router

Rin rintrum at gmail.com
Fri Oct 1 05:34:08 EDT 2010


Hi group, 

I have a 7609 PE router with 2 TenGigabit interfaces uplink to other P
routers and Gigabit interfaces downlink to access switch. Recently, I detect
the CPU of that router is punted approximately every 35mins and result in
below output:
==============================================================  
PE-Router#sho proc cpu hist
 

    6666666644444555552222211111333331111133333111115555511111
100                                                           
 90                                                           
 80                                                           
 70                                                           
 60                                                           
 50                                                           
 40                                                           
 30                                                           
 20                                                           
 10 ********     *****                              *****     
   0....5....1....1....2....2....3....3....4....4....5....5....
             0    5    0    5    0    5    0    5    0    5    
               CPU% per second (last 60 seconds)
                                                              
     88 6                                    99               
    5664764335455448544554465856444344555656540444344454357766
100                                                           
 90  **                                      **               
 80  **                                      **               
 70  ** *                                    **               
 60  ** *                                    **               
 50  ** *                                    **               
 40  #* *                                    **               
 30  #* *                                    #*               
 20  #* *                                    #*               
 10 *## #*   * **  **  **  *****      *******##       *  *****
   0....5....1....1....2....2....3....3....4....4....5....5....
             0    5    0    5    0    5    0    5    0    5    
               CPU% per minute (last 60 minutes)
              * = maximum CPU%   # = average CPU%
                                                                          
    989968999998999999999979999988898989999887299889 8888 9999988989884889
    7653706465425842251504506532186452964008644208308676970312489352156780
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%
==============================================================

I can confirm this caused by interrupt by configure ERM feature to detect
the interrupt on the router and check CPU right after the interrupt happens.
The following output display the logging: 
============================================================================
===
*Sep 29 13:39:38.830: %SYS-4-CPURESRISING: System is seeing global cpu util
97% at total level more than the configured critical limit 80 %
*Sep 29 13:39:38.830: %SYS-4-CPURESRISING: System is seeing global cpu util
97% at interrupt level more than the configured critical limit 70 %
*Sep 29 13:39:48.830: %SYS-6-CPURESFALLING: System is no longer seeing
global high cpu at total level for the configured critical limit 80%,
current value 44%
*Sep 29 13:39:48.830: %SYS-6-CPURESFALLING: System is no longer seeing
global high cpu at interrupt level for the configured critical limit 70%,
current value 44%
*Sep 29 13:42:13.750: %SYS-5-CONFIG_I: Configured from console by admin on
vty0 (172.16.252.251)
*Sep 29 14:14:11.122: %SYS-5-CONFIG_I: Configured from console by admin on
vty0 (172.16.252.251)
*Sep 29 14:16:58.862: %SYS-4-CPURESRISING: System is seeing global cpu util
90% at total level more than the configured critical limit 80 %
*Sep 29 14:16:58.862: %SYS-4-CPURESRISING: System is seeing global cpu util
90% at interrupt level more than the configured critical limit 70 %
*Sep 29 14:17:08.862: %SYS-6-CPURESFALLING: System is no longer seeing
global high cpu at total level for the configured critical limit 80%,
current value 70%
*Sep 29 14:17:13.862: %SYS-6-CPURESFALLING: System is no longer seeing
global high cpu at interrupt level for the configured critical limit 70%,
current value 37%
*Sep 29 14:22:18.046: %SYS-5-CONFIG_I: Configured from console by admin on
vty0 (172.16.252.251)
*Sep 29 14:56:58.886: %SYS-4-CPURESRISING: System is seeing global cpu util
86% at total level more than the configured critical limit 80 %
*Sep 29 14:56:58.886: %SYS-4-CPURESRISING: System is seeing global cpu util
86% at interrupt level more than the configured critical limit 70 %
*Sep 29 14:57:08.886: %SYS-6-CPURESFALLING: System is no longer seeing
global high cpu at total level for the configured critical limit 80%,
current value 78%
*Sep 29 14:57:13.886: %SYS-6-CPURESFALLING: System is no longer seeing
global high cpu at interrupt level for the configured critical limit 70%,
current value 44%
============================================================================
===

============================================================================
===
PE-Router#sho proc cpu sort 5min 
CPU utilization for five seconds: 1%/1%; one minute: 3%; five minutes: 6%
 PID Runtime(ms)   Invoked      uSecs   5Sec   1Min   5Min TTY Process 
 436   1566429721118606284        140  0.00%  0.35%  0.35%   0 Port manager
per 
   7   205461504  12733776      16135  0.00%  0.33%  0.32%   0 Check heaps

 433       12936     87590        147  0.15%  0.01%  0.18%   1 Virtual Exec

 334    93262848  29028603       3212  0.00%  0.14%  0.14%   0 HIDDEN VLAN
Proc 
 200    40238324 301738465        133  0.00%  0.13%  0.13%   0 IP Input

 202     2747572  19557286        140  0.00%  0.02%  0.04%   0 ADJ resolve
proc 
  55    31762204    999074      31791  0.00%  0.03%  0.03%   0 Per-minute
Jobs  
 289    25288024  27364507        924  0.00%  0.03%  0.03%   0 IPC LC
Message H 
 486    11390424 313585288         36  0.00%  0.03%  0.01%   0 TDP Hello

  25    20877464  86502304        241  0.00%  0.03%  0.01%   0 IPC Seat
Manager 
 104    11541792  46704022        247  0.00%  0.01%  0.01%   0
esw_vlan_stat_pr 
 313    16750072  85168825        196  0.00%  0.02%  0.00%   0 CEF: IPv4
proces 
============================================================================
===

I followed the troubleshooting high CPU Utilization Guide but still cannot
detect which kind of traffic is punted to CPU. Can anyone in the list show
me the way how to detect which packets sending to CPU on 7609 router? 

Here are the troubleshooting guide that I have checked:
http://www.cisco.com/en/US/products/hw/routers/ps368/prod_tech_notes_list.ht
ml#anchor4
http://www.cisco.com/en/US/products/hw/switches/ps708/products_tech_note0918
6a00804916e0.shtml#utilities

Thanks,
Rin





More information about the cisco-nsp mailing list