Re: hate to harp on SET SHARE but...

2007-08-10 Thread barton
Stephen, i think you missed the fact that she does have a performance monitor that shows 
all this information in great detail.









Stephen Frazier wrote:

You have many users on your system. Why do you think WVLNX2 is getting 
more than 50% of one IFL? Maybe it is getting 50% and WVLNX4 is also 
getting 50% that would cause the HMC System Activity display to show 
about 100%. Maybe WVLNX4, WVLNX5, WVLNX6, and WVLNX8 are all getting 
about 25% each and WVLNX2 is getting about 1%. Lots of other 
possibilities. What evidence do you have that it is WVLNX2 that is the 
problem?


[EMAIL PROTECTED] wrote:


On the HMC, we have a System Activity display that shows our 2 CPs and
the 1 SP.  Normally the IFL is really low, but occassionally, like right
now, it shows around 100% usage.

Anne D. Crabtree
System Programmer
WV Dept of Administration - OT
304-558-1494
Fax 304-558-1441


[EMAIL PROTECTED] 8/8/2007 9:53 AM >>>


What make you think that WVLNX2 is getting more than 50% of your one
IFL?





Re: hate to harp on SET SHARE but...

2007-08-08 Thread Anne Crabtree
In original post, the figures reported on ESALNXP for CPU percent showed
95.3 percent for wvlnx2 at 8:39 this morning...

Anne D. Crabtree
System Programmer
WV Dept of Administration - OT
304-558-1494
Fax 304-558-1441

>>> [EMAIL PROTECTED] 8/8/2007 10:19 AM >>>
You have many users on your system. Why do you think WVLNX2 is getting
more than 50% of one IFL? 
Maybe it is getting 50% and WVLNX4 is also getting 50% that would cause
the HMC System Activity 
display to show about 100%. Maybe WVLNX4, WVLNX5, WVLNX6, and WVLNX8
are all getting about 25% each 
and WVLNX2 is getting about 1%. Lots of other possibilities. What
evidence do you have that it is 
WVLNX2 that is the problem?

[EMAIL PROTECTED] wrote:
> On the HMC, we have a System Activity display that shows our 2 CPs
and
> the 1 SP.  Normally the IFL is really low, but occassionally, like
right
> now, it shows around 100% usage.
> 
> Anne D. Crabtree
> System Programmer
> WV Dept of Administration - OT
> 304-558-1494
> Fax 304-558-1441
> 
 [EMAIL PROTECTED] 8/8/2007 9:53 AM >>>
> What make you think that WVLNX2 is getting more than 50% of your one
> IFL?
> 

-- 
Stephen Frazier
Information Technology Unit
Oklahoma Department of Corrections
3400 Martin Luther King
Oklahoma City, Ok, 73111-4298
Tel.: (405) 425-2549
Fax: (405) 425-2554
Pager: (405) 690-1828
email:  stevef%doc.state.ok.us


Re: hate to harp on SET SHARE but...

2007-08-08 Thread Stephen Frazier
You have many users on your system. Why do you think WVLNX2 is getting more than 50% of one IFL? 
Maybe it is getting 50% and WVLNX4 is also getting 50% that would cause the HMC System Activity 
display to show about 100%. Maybe WVLNX4, WVLNX5, WVLNX6, and WVLNX8 are all getting about 25% each 
and WVLNX2 is getting about 1%. Lots of other possibilities. What evidence do you have that it is 
WVLNX2 that is the problem?


[EMAIL PROTECTED] wrote:

On the HMC, we have a System Activity display that shows our 2 CPs and
the 1 SP.  Normally the IFL is really low, but occassionally, like right
now, it shows around 100% usage.

Anne D. Crabtree
System Programmer
WV Dept of Administration - OT
304-558-1494
Fax 304-558-1441


[EMAIL PROTECTED] 8/8/2007 9:53 AM >>>

What make you think that WVLNX2 is getting more than 50% of your one
IFL?



--
Stephen Frazier
Information Technology Unit
Oklahoma Department of Corrections
3400 Martin Luther King
Oklahoma City, Ok, 73111-4298
Tel.: (405) 425-2549
Fax: (405) 425-2554
Pager: (405) 690-1828
email:  stevef%doc.state.ok.us


Re: hate to harp on SET SHARE but...

2007-08-08 Thread Anne Crabtree
On the HMC, we have a System Activity display that shows our 2 CPs and
the 1 SP.  Normally the IFL is really low, but occassionally, like right
now, it shows around 100% usage.

Anne D. Crabtree
System Programmer
WV Dept of Administration - OT
304-558-1494
Fax 304-558-1441

>>> [EMAIL PROTECTED] 8/8/2007 9:53 AM >>>
What make you think that WVLNX2 is getting more than 50% of your one
IFL?

[EMAIL PROTECTED] wrote:
> On ESAUSRC screen of ESAMON, it shows this:
>   
>   <-SHARE--->
 
> 
>   Account  ACI Grp <-Maximum>   # of
VM
>   Stor
> UserID   ClassID  Code NameRel Abs Type Share Limit CPUs
> Mode Mode
>     -- ---  - - 
>  
> VMSERVU  *Linuxen 1   1500   .  Rel   0.0  Soft1

> XC  V=V 
> WVLNX2   *Linuxen WVLNX2   100   .  Abs  50.0  Hard1

> ESA V=V 
> WVLNX4   *Linuxen WVLNX4   100   .  Rel   0.0  Soft1

> ESA V=V 
> WVLNX5   *Linuxen WVLNX5   100   .  Rel   0.0  Soft1

> ESA V=V 
> WVLNX6   *Linuxen WVLNX6   100   .  Rel   0.0  Soft1

> ESA V=V 
> WVLNX8   *Linuxen WVLNX8   100   .  Rel   0.0  Soft1

> ESA V=V 
>
-
> 
> ESALNXP screen shows this:
> 
> Screen: ESALNXP  State of West Virginia ESAMON 3.7.0 08/08
> 08:26-08:41 
> 1 of 3  LINUX VSI Process Statistics Report NODE * LIMIT 500
> 2086 3E1FE
>  
 
>
> <-Process Ident-> <-CPU
Percents->
> nice
> Time Node Name  IDPPID   GRP   Tot  sys user syst
usrt
> valu
>   - - - -    

> 
> 08:39:00 wvlnx2   oracle19524 1 19524  0.4  0.4  0.1  0.0 
0.0 
>   0
>   oracle19520 1 19520  0.4  0.3  0.1  0.0 
0.0 
>   0
>   oracle19518 1 19518  0.2  0.1  0.0  0.0 
0.0 
>   0
>   oracle19501 1 19501 20.2 15.9  4.3  0.0 
0.0 
>   0
>   oracle19499 1 19499 20.7 17.1  3.6  0.0 
0.0 
>   0
>   oracle19485 1 19485 27.1 21.6  5.5  0.0 
0.0 
>   0
>   oracle19478 1 19478  1.8  1.4  0.5  0.0 
0.0 
>   0
>   oracle19367 1 19367  6.1  4.5  1.7  0.0 
0.0 
>   0
>   oracle19336 1 19336  1.1  0.8  0.3  0.0 
0.0 
>   0
>   snmpd  5460 1  5459  0.5  0.2  0.3  0.0 
0.0 
>  10
>   cron570 1   570  0.3  0.0  0.0  0.3 
0.0 
>   0
>   tnslsnr 505 1   505  1.0  0.6  0.0  0.3 
0.0 
>   0
>   *Totals*  0 0 0 95.3 74.4 20.2  0.6 
0.0 
>   0
> 
> My understanding is that the LIMITHARD should prevent it from taking
> more than 50% of my one IFL.  (We only have 1!)  Did I misunderstand?

> 
> Anne D. Crabtree
> System Programmer
> WV Dept of Administration - OT
> 304-558-1494
> Fax 304-558-1441

-- 
Stephen Frazier
Information Technology Unit
Oklahoma Department of Corrections
3400 Martin Luther King
Oklahoma City, Ok, 73111-4298
Tel.: (405) 425-2549
Fax: (405) 425-2554
Pager: (405) 690-1828
email:  stevef%doc.state.ok.us


Re: hate to harp on SET SHARE but...

2007-08-08 Thread Rob van der Heij
On 8/8/07, Stephen Frazier <[EMAIL PROTECTED]> wrote:

> What make you think that WVLNX2 is getting more than 50% of your one IFL?

Actually, Linux thinks it gets more than 50% of a CPU. Adding up all
the per-process usage that Linux reports, we get 95.3% of a CPU in
this case. This demonstrates nicely that the CPU usage that Linux
itself reports is not correct unless corrected for virtualization
effects.

I followed up with Anne outside the list and identified a
configuration problem that prevented ESALPS from correcting the Linux
numbers. With that correction, the per-process usage reported by
ESALPS is correct.

Rob
-- 
Rob van der Heij
Velocity Software, Inc
http://velocitysoftware.com/


Re: hate to harp on SET SHARE but...

2007-08-08 Thread Stephen Frazier

What make you think that WVLNX2 is getting more than 50% of your one IFL?

[EMAIL PROTECTED] wrote:

On ESAUSRC screen of ESAMON, it shows this:
  
  <-SHARE--->  

  Account  ACI Grp <-Maximum>   # of VM

  Stor
UserID   ClassID  Code NameRel Abs Type Share Limit CPUs
Mode Mode
    -- ---  - - 
 
VMSERVU  *Linuxen 1   1500   .  Rel   0.0  Soft1 
XC  V=V 
WVLNX2   *Linuxen WVLNX2   100   .  Abs  50.0  Hard1 
ESA V=V 
WVLNX4   *Linuxen WVLNX4   100   .  Rel   0.0  Soft1 
ESA V=V 
WVLNX5   *Linuxen WVLNX5   100   .  Rel   0.0  Soft1 
ESA V=V 
WVLNX6   *Linuxen WVLNX6   100   .  Rel   0.0  Soft1 
ESA V=V 
WVLNX8   *Linuxen WVLNX8   100   .  Rel   0.0  Soft1 
ESA V=V 
-


ESALNXP screen shows this:

Screen: ESALNXP  State of West Virginia ESAMON 3.7.0 08/08
08:26-08:41 
1 of 3  LINUX VSI Process Statistics Report NODE * LIMIT 500
2086 3E1FE
   
   
<-Process Ident-> <-CPU Percents->
nice
Time Node Name  IDPPID   GRP   Tot  sys user syst usrt
valu
  - - - -     

08:39:00 wvlnx2   oracle19524 1 19524  0.4  0.4  0.1  0.0  0.0 
  0
  oracle19520 1 19520  0.4  0.3  0.1  0.0  0.0 
  0
  oracle19518 1 19518  0.2  0.1  0.0  0.0  0.0 
  0
  oracle19501 1 19501 20.2 15.9  4.3  0.0  0.0 
  0
  oracle19499 1 19499 20.7 17.1  3.6  0.0  0.0 
  0
  oracle19485 1 19485 27.1 21.6  5.5  0.0  0.0 
  0
  oracle19478 1 19478  1.8  1.4  0.5  0.0  0.0 
  0
  oracle19367 1 19367  6.1  4.5  1.7  0.0  0.0 
  0
  oracle19336 1 19336  1.1  0.8  0.3  0.0  0.0 
  0
  snmpd  5460 1  5459  0.5  0.2  0.3  0.0  0.0 
 10
  cron570 1   570  0.3  0.0  0.0  0.3  0.0 
  0
  tnslsnr 505 1   505  1.0  0.6  0.0  0.3  0.0 
  0
  *Totals*  0 0 0 95.3 74.4 20.2  0.6  0.0 
  0


My understanding is that the LIMITHARD should prevent it from taking
more than 50% of my one IFL.  (We only have 1!)  Did I misunderstand? 


Anne D. Crabtree
System Programmer
WV Dept of Administration - OT
304-558-1494
Fax 304-558-1441


--
Stephen Frazier
Information Technology Unit
Oklahoma Department of Corrections
3400 Martin Luther King
Oklahoma City, Ok, 73111-4298
Tel.: (405) 425-2549
Fax: (405) 425-2554
Pager: (405) 690-1828
email:  stevef%doc.state.ok.us


hate to harp on SET SHARE but...

2007-08-08 Thread Anne Crabtree
On ESAUSRC screen of ESAMON, it shows this:
  
  <-SHARE--->  

  Account  ACI Grp <-Maximum>   # of VM
  Stor
UserID   ClassID  Code NameRel Abs Type Share Limit CPUs
Mode Mode
    -- ---  - - 
 
VMSERVU  *Linuxen 1   1500   .  Rel   0.0  Soft1 
XC  V=V 
WVLNX2   *Linuxen WVLNX2   100   .  Abs  50.0  Hard1 
ESA V=V 
WVLNX4   *Linuxen WVLNX4   100   .  Rel   0.0  Soft1 
ESA V=V 
WVLNX5   *Linuxen WVLNX5   100   .  Rel   0.0  Soft1 
ESA V=V 
WVLNX6   *Linuxen WVLNX6   100   .  Rel   0.0  Soft1 
ESA V=V 
WVLNX8   *Linuxen WVLNX8   100   .  Rel   0.0  Soft1 
ESA V=V 
-

ESALNXP screen shows this:

Screen: ESALNXP  State of West Virginia ESAMON 3.7.0 08/08
08:26-08:41 
1 of 3  LINUX VSI Process Statistics Report NODE * LIMIT 500
2086 3E1FE
   
   
<-Process Ident-> <-CPU Percents->
nice
Time Node Name  IDPPID   GRP   Tot  sys user syst usrt
valu
  - - - -     

08:39:00 wvlnx2   oracle19524 1 19524  0.4  0.4  0.1  0.0  0.0 
  0
  oracle19520 1 19520  0.4  0.3  0.1  0.0  0.0 
  0
  oracle19518 1 19518  0.2  0.1  0.0  0.0  0.0 
  0
  oracle19501 1 19501 20.2 15.9  4.3  0.0  0.0 
  0
  oracle19499 1 19499 20.7 17.1  3.6  0.0  0.0 
  0
  oracle19485 1 19485 27.1 21.6  5.5  0.0  0.0 
  0
  oracle19478 1 19478  1.8  1.4  0.5  0.0  0.0 
  0
  oracle19367 1 19367  6.1  4.5  1.7  0.0  0.0 
  0
  oracle19336 1 19336  1.1  0.8  0.3  0.0  0.0 
  0
  snmpd  5460 1  5459  0.5  0.2  0.3  0.0  0.0 
 10
  cron570 1   570  0.3  0.0  0.0  0.3  0.0 
  0
  tnslsnr 505 1   505  1.0  0.6  0.0  0.3  0.0 
  0
  *Totals*  0 0 0 95.3 74.4 20.2  0.6  0.0 
  0

My understanding is that the LIMITHARD should prevent it from taking
more than 50% of my one IFL.  (We only have 1!)  Did I misunderstand? 

Anne D. Crabtree
System Programmer
WV Dept of Administration - OT
304-558-1494
Fax 304-558-1441