how can i monitor cpu ultilization on MSC

fly · ‎01-06-2014

such as attach and show process cpu command on GSR

thank you!

Tom

mivens · ‎01-07-2014

From the CLI:

"show process cpu location x/y/CPU0"

or

"monitor process location x/y/CPU0"

Or SNMP poll for cpmCPUTotal5secRev, cpmCPUTotal1minRev or cpmCPUTotal5minRev from the CISCO-PROCESS-MIB. For example for an MSC in slot 0, the 1 min CPU average:

CISCO-PROCESS-MIB::cpmCPUTotal1minRev.2 = Gauge32: 2 percent

CISCO-PROCESS-MIB::cpmCPUTotalPhysicalIndex.2 = INTEGER: 2359704

ENTITY-MIB::entPhysicalName.2359704 = STRING: 0/0/* - cpu

ENTITY-MIB::entPhysicalContainedIn.2359704 = INTEGER: 38645760

ENTITY-MIB::entPhysicalDescr.38645760 = STRING: Cisco CRS-1 Series Modular Services Card

Rivalino Tamaela · ‎01-07-2014

Hello,

If you want to attach to specific card for monitoring, you can do as follows to monitor CPU consumer:

RP/0/RP1/CPU0:CRS-C#show platform

Tue Jan 7 09:08:47.213 EDT

Node Type PLIM State Config State

------------- ----------------- ---------------- --------------- ---------------

0/0/CPU0 MSC-B 8-10GbE IOS XR RUN PWR,NSHUT,MON

0/1/CPU0 FP40 4-10GbE IOS XR RUN PWR,NSHUT,MON

0/2/CPU0 MSC-B Jacket Card IOS XR RUN PWR,NSHUT,MON

0/3/CPU0 MSC 16OC48-POS/DPT IOS XR RUN PWR,NSHUT,MON

0/5/CPU0 MSC 1OC768-POS IOS XR RUN PWR,NSHUT,MON

0/RP0/CPU0 RP(Standby) N/A IOS XR RUN PWR,NSHUT,MON

0/RP1/CPU0 RP(Active) N/A IOS XR RUN PWR,NSHUT,MON

RP/0/RP1/CPU0:CRS-C#run attach 0/5/cpu0

Tue Jan 7 09:08:54.964 EDT

attach: Starting session 2 to node 0/5/cpu0

# top_procs -D

Computing times...

node0_5_CPU0: 143 procs, 1 cpu, 1.04 delta, 858:02:56 uptime

Memory: 2048 MB total, 928.292 MB free, sample time: Tue Jan 7 09:09:02 2014

cpu 0 idle: 97.31%, kernel: 0.09%

pid mem MB user cpu kernel cpu delta % ker % tot name

73796 1.203 46671.271 82.590 0.014 0.00 1.49 pse_driver

90192 0.367 14455.991 3385.648 0.008 0.19 0.79 stats_svr

3117200 0.234 0.016 0.001 0.002 0.00 0.19 top_procs

73798 0.804 1271.587 360.618 0.001 0.09 0.09 fab_svr

12296 0.308 0.016 0.004 0.001 0.00 0.09 devc-pty

86060 0.308 4.165 3.190 0.000 0.00 0.00 diag_svr

41003 3.648 1715.777 469.277 0.000 0.00 0.00 wdsysmon

73799 3.792 1107.399 222.178 0.000 0.00 0.00 gsp

65590 0.328 98.651 26.463 0.000 0.00 0.00 fabio_svr

12290 0.242 0.613 0.671 0.000 0.00 0.00 wd-critical-mon

node0_5_CPU0: 143 procs, 1 cpu, 5.13 delta, 858:03:01 uptime

Memory: 2048 MB total, 928.292 MB free, sample time: Tue Jan 7 09:09:07 2014

cpu 0 idle: 97.28%, kernel: 0.01%

pid mem MB user cpu kernel cpu delta % ker % tot name

73796 1.203 46671.347 82.590 0.075 0.00 1.51 pse_driver

90192 0.367 14456.019 3385.653 0.033 0.09 0.65 stats_svr

73798 0.804 1271.592 360.619 0.006 0.01 0.11 fab_svr

90191 2.472 3372.803 30.848 0.005 0.00 0.09 egressq

73799 3.792 1107.402 222.178 0.003 0.00 0.05 gsp

3117200 0.234 0.018 0.002 0.003 0.01 0.05 top_procs

110687 3.000 123.548 69.889 0.003 0.01 0.05 fib_mgr

241774 4.031 227.568 16.354 0.002 0.00 0.03 qos_ea

65593 2.265 822.075 75.675 0.001 0.00 0.01 fabricq_mgr

110675 5.988 16.101 5.816 0.001 0.00 0.01 netio

# exit

Terminating attach to node 0/5/cpu0..

RP/0/RP1/CPU0:CRS-C#

And the following is to monitor memory consumer from top to bottom:

RP/0/RP1/CPU0:CRS-C#run attach 0/5/cpu0

Tue Jan 7 09:11:52.798 EDT

attach: Starting session 2 to node 0/5/cpu0

# top_procs -MD

Computing memory...

node0_5_CPU0: 143 procs, 1 cpu, 1.06 delta, 858:05:53 uptime

Memory: 2048 MB total, 928.292 MB free, sample time: Tue Jan 7 09:11:59 2014

cpu 0 idle: 96.52%, kernel: 0.69%

pid mem MB user cpu kernel cpu delta % ker % tot name

110675 5.988 16.102 5.816 0.000 0.00 0.00 netio

8195 5.796 11.635 1.200 0.000 0.00 0.00 dllmgr

241774 4.031 227.581 16.355 0.000 0.00 0.00 qos_ea

73799 3.792 1107.451 222.191 0.000 0.00 0.00 gsp

41003 3.648 1715.886 469.302 0.000 0.00 0.00 wdsysmon

65595 3.121 1.305 0.771 0.000 0.00 0.00 fia_driver

110687 3.000 123.554 69.896 0.000 0.00 0.00 fib_mgr

110678 2.847 62.335 29.719 0.000 0.00 0.00 platform_mgr

241772 2.738 2.903 2.091 0.000 0.00 0.00 l2fib_mgr

110698 2.652 3.917 1.337 0.000 0.00 0.00 tcam_mgr

node0_5_CPU0: 143 procs, 1 cpu, 5.15 delta, 858:05:58 uptime

Memory: 2048 MB total, 928.292 MB free, sample time: Tue Jan 7 09:12:04 2014

cpu 0 idle: 97.33%, kernel: 0.03%

pid mem MB user cpu kernel cpu delta % ker % tot name

110675 5.988 16.102 5.816 0.000 0.00 0.00 netio

8195 5.796 11.635 1.200 0.000 0.00 0.00 dllmgr

241774 4.031 227.581 16.355 0.000 0.00 0.00 qos_ea

73799 3.792 1107.456 222.192 0.006 0.01 0.11 gsp

41003 3.648 1715.886 469.302 0.000 0.00 0.00 wdsysmon

65595 3.121 1.305 0.771 0.000 0.00 0.00 fia_driver

110687 3.000 123.554 69.897 0.001 0.01 0.01 fib_mgr

110678 2.847 62.336 29.719 0.001 0.00 0.01 platform_mgr

241772 2.738 2.903 2.091 0.000 0.00 0.00 l2fib_mgr

110698 2.652 3.917 1.337 0.000 0.00 0.00 tcam_mgr

#

Regards,

Rivalino

dpothier · ‎01-07-2014

A word of caution here, "monitor process location r/s/m" will actually drive the CPU utilization up on the given LC. Also note if you are monitoring resources, you will also want to monitor the forwarding ASIC for a given LC via "show controller pse utilization {ingress/egress} location r/s/m".