100% Utilization in Server

[Former Member]

Hi,

Application Server UNX0124 on QR1( Quality ECC Ssytem)doesnu2019t seem right. Constantly shown as 100% utilisation, memory usage the same etc, even over wekeend when it didnu2019t seem that busy with jobs. Not sure if theres a problem with stats collector or some rogue process is using all CPU.Can you please suggest me.

QR1 is the dialog Instance running on AIX 6.1 & Oracle 10.2.

[Former Member · Accepted solution]

One more thing I have checked the OS Collector availavle days-- It is showing not available but when I am checking from AIX level with saposcol -s command :

Collector Versions :

running : COLL 20.95 640 - AIX v11.16 5L-64 bit 080318

dialog : COLL 20.95 640 - AIX v11.16 5L-64 bit 080318

Shared Memory : attached

Number of records : 17640

Active Flag : active (01)

Operating System : AIX unxa0016 3 5 00E6653B4C00

Collector PID : 327726 (0005002E)

Collector : running

Start time coll. : Tue Jun 7 16:41:40 2011

Current Time : Tue Jul 5 15:13:54 2011

Last write access : Sat Jun 18 21:18:57 2011

Last Read Access : Tue Jul 5 15:13:35 2011

Collection Interval : 10 sec (next delay).

Collection Interval : 10 sec (last ).

Status : free

Collect Details : required

Refresh : required

Header Extention Structure

Number of x-header Records : 1

Number of Communication Records : 60

Number of free Com. Records : 60

Resulting offset to 1.data rec. : 61

Trace level : 2

Collector in IDLE - mode ? : NO

become idle after 300 sec without read access.

Length of Idle Interval : 60 sec

Length of norm.Interval : 10 sec

Please help me.

[Former Member · Accepted solution]

Hello Das,

Can you go to OS07n transaction and see the latest data collection time stamp.

If its not consistent then restart the OSCOL once.

Also you can compare the data seen in OS07n and operating system level data too.

Regards

Vivek

[Former Member · Accepted solution]

Vivek I have stopped the collector by saposcol -k. But when I am checking the status of the collector it is still showing running.The issue is still there even after restarting it.One thing I noticed from saposcol -s command :

Last write access : Sat Jun 18 21:18:57 2011..

[Former Member · Accepted solution]

Hi

I even tried to run saposcol in the dialog mode but still no result.

saposcol -d

*****************************************************************************

• This is Saposcol Version COLL 20.95 640 - AIX v11.16 5L-64 bit 080318

• Please use 'help' to see the usage.

*****************************************************************************

Collector > help

help

SAPOSCOL Command List:

stat - show Collector Status and Shared Memory information

launch - start Collection Process

kill - stop Collection Process

version - show Collector Version

quit - stop this dialog program (not the collecting process)

force - force collector start in background

killprint - stop collecting process and print data

interval - modify collecting interval

parameters: <interval [sec]>

dump - print shared memory data:

parameters: <type> [<subtype>]

type = hour | cpu | memory | disk | lan

| top | config | config2 | proc | net

| strpt | virt_config | virt_metrics

subtype = single | sum | all | used | config

clean - initialize the shared memory

leave - delete shared memory and detach

getidle - time when collector enters 'idle' state if there is no read access

show current value when called without a value <sec>

can only be set before the collector starts

idleint - length of an 'idle' interval

show current value when called without a value <sec>

can only be set before the collector starts.

ask - send a question to running saposcol

Collector > kill

kill

Setting Stop Flag :

11:40:21 06.07.2011 LOG: ==== Stop Flag was set by saposcol (kill_collector()).

11:40:21 06.07.2011 LOG: ==== The collection process will stop as soon as possible

********************

can't kill process 327726.

kill: No such process

ERROR:No reaction from collecting process 327726.

Please kill collecting process.

Collector > kill -9 327726

kill -9 327726

Setting Stop Flag :

11:41:57 06.07.2011 LOG: ==== Stop Flag was set by saposcol (kill_collector()).

11:41:57 06.07.2011 LOG: ==== The collection process will stop as soon as possible

********************

can't kill process 327726.

kill: No such processERROR:No reaction from collecting process 327726.

Please kill collecting process.

Collector > launch

launch

11:43:38 06.07.2011 LOG: Effective User Id is root

*************************************************************************

• The OS Collector (PID 327726) is already running .....

[former_member185031 · Accepted solution]

Identify the process through ps -ef | grep sap and than use kill -9

Regards,

Subhash

[Former Member · Accepted solution]

Subhash I tried to kill the process as you said but not able to do it.Check the log.

saposcol -l

-l

12:21:17 06.07.2011 LOG: Effective User Id is root

***********************************************************************

• This is Saposcol Version COLL 20.95 640 - AIX v11.16 5L-64 bit 080318

• Usage: saposcol -l: Start OS Collector

• saposcol -k: Stop OS Collector

• saposcol -d: OS Collector Dialog Mode

• saposcol -s: OS Collector Status

• The OS Collector (PID 327726) is already running .....

************************************************************************

kill -9 327726

kill: 327726: 0403-003 The specified process does not exist.

[Former Member · Accepted solution]

Hello Das,

Go to saposcol -d

Then kill the process use clean command.

Then force the start of OScol.

This should fix it.

If not directly force the start of the process which generally kills the older process and restarts the process with new process ID.

Regards

Vivek

[Former Member · Accepted solution]

Spot on Sir..It resolved.