Mailing List Archive: Useful exception processor statistics?

Useful exception processor statistics?

Feb 11, 2003, 2:53 PM

Post #1 of 3 (1236 views)

Are there any ways to do some useful analysis of what is sucking up
exception processor cpu?

SBR(xxx.xxxx vty)# sh sched
Total uptime 8+23:23:35, (775415045 ms), 478792393 thread dispatches
CPU load is 66% (5 second), 59% (1 minute)
Total network interrupt time 85054125386 (usec)

CPU Name Time(ms)
77% Idle 602664323
7% Threads 54398637
15% ISR 118352085
14% Level 1 115670085
0% Level 5 2682000

The cpu load and idle loads seem to be... Misleading... Is it just that
the breakdowns are the average of the entire uptime of the device, or am I
missing something?

--
Richard A Steenbergen <ras@e-gerbil.net> http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)

Useful exception processor statistics? [ In reply to ]

rkjnsp at ieg

Feb 11, 2003, 5:17 PM

Post #2 of 3 (1213 views)

Permalink

Indirect analysis ? I would try turning off SNMP, filtering/counting packets
that are known to generate exceptions...

Rubens

----- Original Message -----
From: "Richard A Steenbergen" <ras@e-gerbil.net>
To: <juniper-nsp@puck.nether.net>
Sent: Tuesday, February 11, 2003 5:54 PM
Subject: [j-nsp] Useful exception processor statistics?

| Are there any ways to do some useful analysis of what is sucking up
| exception processor cpu?
|
| SBR(xxx.xxxx vty)# sh sched
| Total uptime 8+23:23:35, (775415045 ms), 478792393 thread dispatches
| CPU load is 66% (5 second), 59% (1 minute)
| Total network interrupt time 85054125386 (usec)
|
| CPU Name Time(ms)
| 77% Idle 602664323
| 7% Threads 54398637
| 15% ISR 118352085
| 14% Level 1 115670085
| 0% Level 5 2682000
|
| The cpu load and idle loads seem to be... Misleading... Is it just that
| the breakdowns are the average of the entire uptime of the device, or am I
| missing something?
|
| --
| Richard A Steenbergen <ras@e-gerbil.net> http://www.e-gerbil.net/ras
| GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)
| _______________________________________________
| juniper-nsp mailing list juniper-nsp@puck.nether.net
| http://puck.nether.net/mailman/listinfo/juniper-nsp

Useful exception processor statistics? [ In reply to ]

josefb at juniper

Feb 12, 2003, 4:12 AM

Post #3 of 3 (1219 views)

Permalink

At 08:54 PM 2/11/2003, Richard A Steenbergen wrote:
>Are there any ways to do some useful analysis of what is sucking up
>exception processor cpu?
>
>SBR(xxx.xxxx vty)# sh sched
>Total uptime 8+23:23:35, (775415045 ms), 478792393 thread dispatches
>CPU load is 66% (5 second), 59% (1 minute)
>Total network interrupt time 85054125386 (usec)
>
> CPU Name Time(ms)
> 77% Idle 602664323
> 7% Threads 54398637
> 15% ISR 118352085
> 14% Level 1 115670085
> 0% Level 5 2682000
>
>The cpu load and idle loads seem to be... Misleading...

Richard,
those commands are debug commands and not intended
to be used for customers .....

The values under CPU are presented against the total
uptime and therefor the difference in the values of
CPU load for 5 seconds and one minute. In order to find
out where we use the CPU load I would like to encourage
you to open a case to the support alias As mentioned already
in another note sampling is a common application but
it could be other stuff too ...

thanks
Josef

> Is it just that
>the breakdowns are the average of the entire uptime of the device, or am I
>missing something?
>
>--
>Richard A Steenbergen <ras@e-gerbil.net> http://www.e-gerbil.net/ras
>GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1 2CBC)
>_______________________________________________
>juniper-nsp mailing list juniper-nsp@puck.nether.net
>http://puck.nether.net/mailman/listinfo/juniper-nsp