Resource consumption in zones

Let's say I have a beefy x86 box with 4 CPUs and 32Gb of RAM. Let's also say that I set up 30-50 zones on this machine, each one running a different web application (or sets of web applications), or just doing generic web serving.
What are the constraints around TCP/IP load in this scenario (e.g where is the first limit I'll hit)? Will there be significant performance degradation as the zones start consuming huge numbers of TCP connections each?
What about total number of processes? Am I limited to the process limit of the global zone?
Thanks,
-chris

Zones themselves will not add any additional constraints in this
situation. Think of them as a partitioning mechanism - if you had
run the same workloads in just the global zone, there would be
some characterization of the TCP/IP load on the system. Placing
those workloads into individual zones will not increase the load
on the system - rather, you've split the existing workload into
separate pieces, so to speak.

Similar Messages

Resource consumption details not shown for each row source operation

Hi Friends,
I have generated a tracefile for a query as shown below.
set autotrace traceonly arraysize 100
ALTER SESSION SET optimizer_mode = ALL_ROWS ;
alter session set tracefile_identifier = 'trcwithtime_221010_03';
alter session set events '10046 trace name context forever, level 12';
spool on
spool C:\Sree\trcwtime.log
-- put your statement here
SELECT e.*,d.dname
FROM scott.emp e
    ,scott.dept d
WHERE e.deptno = d.deptno;
spool off;
exitBelow is the trace file
SELECT e.*,d.dname
FROM scott.emp e
    ,scott.dept d
WHERE e.deptno = d.deptno
call     count       cpu    elapsed       disk      query    current        rows
Parse        1      0.00       0.00          0          0          0           0
Execute      1      0.00       0.00          0          0          0           0
Fetch        2      0.00       0.00          4          7          0          14
total        4      0.00       0.00          4          7          0          14
Misses in library cache during parse: 0
Optimizer goal: ALL_ROWS
Parsing user id: 173
Rows     Row Source Operation
     14 HASH JOIN
      4   TABLE ACCESS FULL DEPT
     14   TABLE ACCESS FULL EMP
Elapsed times include waiting on following events:
Event waited on                             Times   Max. Wait Total Waited
----------------------------------------   Waited ---------- ------------
SQL*Net message to client                       2        0.00          0.00
db file sequential read                         4        0.00          0.00
SQL*Net message from client                     2        0.41          0.77
********************************************************************************For each rowsource operation, resource consumption details (cr,pr,pw,time) are not shown in the TKPROF output.
How can i get these details as well?
DBVersion: 9.2.0.8
Regards,
Sreekanth Munagala

Sreekanth Munagala wrote:
For each rowsource operation, resource consumption details (cr,pr,pw,time) are not shown in the TKPROF output.
How can i get these details as well?
DBVersion: 9.2.0.8You can get those details by setting values for following parameters (at session level, preferably) before generating the 10046 trace.
ALTER SESSION SET statistics_level = ALL ;
ALTER SESSION SET timed_statistics = TRUE ;

Resource consumption in different bucket

Hi all,
we have some planned orders which are created by optimizer and they
have start date and end date are in different bucket then the total
resource consumption is showed in the bucket where start date of the
order fall. that means actually there is no over consumption of
resource and production is falling under next bucket as per end date.
When we observe the same in capacity planning book its showing in some
buckets as capacity over loaded and production is showing in next
bucket. Logically it should not the case.
Kindly suggest how to proceed in this senario
Thanks in advance.
Regards

THX

Does hiding controls/indicators on the front panel significantly effect computer resource consumption?

I remember reading something about transparent controls and indicators leading to overhead processing... does hiding controls/indicators have similar or ANY negative effects?

There's a difference between "effecting resource consumption" and "affecting resource consumption" - I assume you mean "affect".
There's a difference between a transparent control and a hidden control - Hidden means the VISIBLE property is set to FALSE, and certain optimizations can be done - it doesn't have to draw anything. If a control is set to transparent, it will still go through the motions of converting that integer to text (or whatever), and then draw nothing (the decision is made later, so more time is taken).
A transparent control can still react to mouse clicks. A hidden one cannot.
It shouldn't affect any memory consumption, except for a video frame buffer.
It's not clear from the question whether you mean the ACT of hiding t
hem (the turning OFF of the visible property), or the fact that they may be hidden.
Steve Bird
Culverson Software - Elegant software that is a pleasure to use.
Culverson.com
Blog for (mostly LabVIEW) programmers: Tips And Tricks

Resource Consumption between WLP 8.1 and 10.3

Hi there,
Is there any document about comparison in term of performance and resource consumption between the same portal application running on version 8 or 10?
Thanks.

No.
In all probability the bottleneck will be your application, not the portal framework, so any test would need to be done for your application.
Weblogic Server is definitely faster and JDK1.5 is way faster than 1.4 so you would see improvements.
regards
deepak

Production Quantity with Resource Consumption - 9ASPROD

Hi ,
Could anyone please tell me how keyfigure 9ASPROD - production(total) which is assigned to keyfigure function 6003 (Production Quantity with Resource Consumption ) is calculated .
We can view this keyfigure in Capcaity check Data view of planning book 9ASNP94.
Thanks in Advance.
Regards,
Srini

Hello Srini,
It calculates in this way: operation end date on production resource +
GR processing time.
The best regards,
Michel Bohn

Fixed PPM/PDS resource consumption and fixed PPM/PDS material consumption

hi experts
while going through help.sap to find out the meaning of linear optimization i came across two phrases ""fixed PPM/PDS resource consumption and fixed PPM/PDS material consumption"" i have worked on SCC03 PPM and curto_simu PDS but i couldn't make out what these phrases means . my perception is that every shop floor operator is given a target to produce certain number of quantities per shift on a production line .Does these phrase means that ?? or it has some other connotation. ??
experts can you please provide clarity on those phrases.what does it actually mean ?

sahu it is even more confusing see "Fixed PPM consumption is the fixed time required independent of quantity produced.""
fixed time of what ?? is it the resource timing or shift timing
Time required to produced material consumption quantity = Fixed cons + variable cons
can i say it is the time required to produce a given lot size quantity ??
sahu would you please put it in very layman terms.

Resource Consumption

Hi,
We are using CTM to create PP/Ds order using PP/DS PPM, we have used Multimixed resource in the PPM definition.
The number of capacities defined is 3 and available time is 24 Hrs, so total available capacity for a day is 72 Hrs.
Now when we are planning for demand of 72 KG with throughput as 1Kg per Hr, the system should schedule the order at the start of the day and complete the order in one day assuming that 100% resource capacity is available.
But when we take the CTM run it is creating a Planned Order which occupies the capacity of one resource for 3 days, i.e. it occupies only 1/3rd capacity and remaining 2/3rd capacity is idle.
The expected result for me is that the system should consider the total available capacity irrespective of the no of capacities defined for the resource and create one planned order even if it occupies multiple capacities for the same resource
Can anyone please suggest how I can consider single capacity for multiple resource of same type using CTM and PP/DS PPM.
Thanks & Regards,
Sanjog

Sanjog,
I have never done this in CTM, but in PP/DS you achieve this with splitting the orders across resources (work centers).
In ECC, in the routing, in the detail of the operation, set the 'required splitting' flag, and then set the number of splits at three (CA02).
Re-cif across (in order to eliminate confusion, it is probably best to delete the old PPM in APO first). Now, in your PPM, you will see in the PPM operation Mode, the Fixed continual consumption is '3' and the variable duration has been reduced to one/third of the original value.
Rgds,
DB49

Freezing problem with resource capping with zones

I've just installed an set up an M4000 box with 4 CPUs and 64Gb of memory. I've installed one zone with CPU and Memory capping as such:
global-mac# zonecfg -z myzone
zonecfg:myzone> info
zonename: myzone
zonepath: /zones/myzone
brand: native
autoboot: true
bootargs:
pool:
limitpriv:
scheduling-class:
ip-type: exclusive
fs:
     dir: /data
     special: /myzfs
     raw not specified
     type: lofs
     options: [nodevice]
net:
     address not specified
     physical: nxge0
     defrouter not specified
capped-cpu:
     [ncpus: 1.00]
capped-memory:
     physical: 16G
     [swap: 8G]
     [locked: 4G]
rctl:
     name: zone.cpu-cap
     value: (priv=privileged,limit=100,action=deny)
rctl:
     name: zone.max-swap
     value: (priv=privileged,limit=8589934592,action=deny)
rctl:
     name: zone.max-locked-memory
     value: (priv=privileged,limit=4294967296,action=deny)
zonecfg:myzone>
Everything seems fine until I get into the zone and move about with the CD command, then everything will freeze for a while and then a few other commands (even a few returns at the prompt) and it will freeze again. Has anyone come across this problem before? Anyone used this sort of resource capping?
If I change the zone configuration to remove the capped resources, the zone works just fine...

<script>
var mymsg = "<bean:message key="image.details" />";
</script>Javascript is just content, so you can dynamically
add your messages to it wherever you want.thanks
gimbel2 but how can we store this in jsp variables.

Experience about resource consumption about migration SQL- server to MaxDB

Hi,
Has somebody experience about migration MS SQL server (running under MS Windows) to MaxDB (Linux). Right now DB size is 2TB and SAP system is used about 1000 SAP users. I need information how cpu- and memory- consumption will change. Otherwise which requirements have to make for target database especially SAPS and hardware design.
Mit freundlichen Grüssen
Andreas Sartor

Check
https://www.sdn.sap.com/irj/sdn/maxdb
--> "Ensuring a Successful Database Migration to SAP MaxDB"
I would book that service.
The people doing that have lots of experience in diagnosing possible problems before the migration and will give you a lot of useful tips.
Markus

ITunes 7 Resource Consumption

Upgraded to iTunes 7 and it is hogging resources. This results in many symptoms that have been reported here already.
My system is unusable while running iTunes 7 after upgrading today from 6.05.
Very weird how one piece of software can mess up your life style.

One single Airport Express (I only own one) and I an just trying to play to the Airport Express.
I have tried selecting multiple speakers but it make no difference, it can't be ticked in the box, even though it is visible.
I have a lot of Mac experience by the way, so I have checked and tried all the usual and unusual things. Even tried turning the Firewall off but no joy.
Ian

Resource consumption for increasing column size

Hello,
DB: 11.2.0.1.0
We need to increase the size of a column in the huge table (~ 400 Gb) from NUMBER(9) to NUMBER(18). Could you tell me what's the internals of the command
alter table foo modify (bar NUMBER (18))Will it change the dictionary only, or it must scan the table block by block to make changes?
Thank you in advance,
Constantine

rp0428 wrote:
>
Thats where i was pointing... existing column values from existing row(s) are checked .. isnt it ...
>
No - that is wrong for OPs use case.
A change from NUMBER(9) to NUMBER(18) is a dictionary-only change.
Existing data DOES NOT need to be checked at all since it is impossible for existing data to conform to NUMBER(9) but not conform to NUMBER(18).Thank you very much rp0428. You're right. Karan just tried to turn my attention to all possible details of problem although I've stated at the begin of discussion that I need to increase the precision.

Zone Resource Control example website

G'Day folks,
I've began work on a website to list examples of setting zone resource controls (processor sets, FSS, ...).
http://www.brendangregg.com/zones.html
Now before I get carried away, does anyone think that maybe I shouldn't - if so I will take it offline no problem.
Just checking I dont get replies like,
"please don't document zones as it's all about to change", or something like that ...
PS. I was pretty happy to see FSS work - that's really cool! :-)
thanks,
Brendan Gregg
[Sydney, Australia]

G'Day Gary,
Thanks for your comments!
Here are some corrections/additions:
The best way to check if pools are enabled is to run
pooladm and examine the output. poold may not be
running and pools may still be active, so performing
a pgrep for poold isn't conclusive.Yes, good point.
In commands 2, you don't need the "final pooladm -c".
The poolcfg -df command will have already done
e this.The final pooladm -c didn't make sense the first time I encountered it, so I had made pretty sure that it was necessary. I'll redo my checks and update the site. Without it, I'm only using poolcfg to make changes, and the man page for poolcfg has:
# man poolcfg
     If you use the -d option,
     the modifications occur to the kernel state. Actual activa-
     tion of the resulting configuration is achieved by way of
     the pooladm(1M) command.It seems to imply a pooladm command is necessary for activation (I used a pooladm -c).
In Resources CPU1 it may be worth mentioning that you
can immediately modify pool bindings for zones if you
use poolbind.
e.g.
/usr/sbin/poolbind -p work1-pool -i zoneid workzone1thanks, I hadn't seen this before!
At the end of Resource CPU 2. It isn't poold which
decides how to allocate CPUs when a configuration is
applied to a system using pooladm -c. This work is
done inside libpool (3LIB) by the pool_conf_commit
API.Ok, I haven't seen documentation on what specifically happens after typing pooladm -c. The documentation such as the poold man page draws a lot of attention to poold:
# man poold
     poold periodically examines the load on the system and
     decides whether intervention is required to maintain optimal
     system performance with respect to resource consumption.
     poold also responds to externally initiated (with respect to
     poold) changes of either resource configuration or objec-
     tives.
     If intervention is required, poold attempts to reallocate
     the available resources to ensure that performance objec-So a change in resource configuration sounds like it means "pooladm -c", and poold is responding.
I'm not sure you meant pool_conf_commit, the man page describes it as writing the config to permanent storage - not triggering an allocation of CPUs.
At my work blog, I have various hints, tips and
tricks for Solaris Resource Pools. If you want to add
a link to my blog (where I try and focus on Solaris
Resource Management issues) feel free. It's at:
http://blogs.sun.com/garypen
Great, I've linked it already, and uploaded a new version (and I imagine I'll upload more versions as I learn more about resource controls).
thanks,
Brendan

Zone Cluster - oracle_server resource create

I am having a problem trying to create an oracle_server resource for my zone cluster.
I have a 2-node zone cluster that utilizes a shared storage zpool to house an Oracle installation and its database files. This is a test system so don't worry about the Oracle setup. I obviously wouldn't put the application and database files on the same storage.
When I run the following command from a global zone member:
clrs create -Z test -g test-rg -t SUNW.oracle_server -p Connect_string=user/password-p ORACLE_SID=SNOW -p ORACLE_HOME=/u01/app/oracle/product/10.2.0/db_snow -p Alert_log_file=/u01/app/oracle/product/10.2.0/db_snow/admin/SNOW/bdump -p Restart_type=RESOURCE_RESTART -p resource_dependencies=test-storage-rs test-oracle_server-rsI get the following errors:
clrs: taloraan.admin.uncw.edu:test - Validation failed. ORACLE_HOME /u01/app/oracle/product/10.2.0/db_snow does not exist
clrs: taloraan.admin.uncw.edu:test - ALERT_LOG_FILE /u01/app/oracle/product/10.2.0/db_snow/admin/SNOW/bdump doesn't exist
clrs: taloraan.admin.uncw.edu:test - PARAMETER_FILE: /u01/app/oracle/product/10.2.0/db_snow/dbs/initSNOW.ora nor server PARAMETER_FILE: /u01/app/oracle/product/10.2.0/db_snow/dbs/spfileSNOW.ora exists
clrs: taloraan.admin.uncw.edu:test - This resource depends on a HAStoragePlus resouce that is not online on this node. Ignoring validation errors.
clrs: tatooine.admin.uncw.edu:test - ALERT_LOG_FILE /u01/app/oracle/product/10.2.0/db_snow/admin/SNOW/bdump doesn't exist
clrs: (C189917) VALIDATE on resource test-oracle_server-rs, resource group test-rg, exited with non-zero exit status.
clrs: (C720144) Validation of resource test-oracle_server-rs in resource group test-rg on node tatooine failed.
clrs: (C891200) Failed to create resource "test:test-oracle_server-rs".So obviously, the crls command cannot find the files (which are located on my shared storage). I am guessing I need to point the command at a global mountpoint.
Regardless, can anyone shed some light on how I make this happen?
I am referencing http://docs.sun.com/app/docs/doc/821-0274/chdiggib?a=view
The section that reads "Example 3 Registering Sun Cluster HA for Oracle to Run in a Zone Cluster"

The storage is mounted but it only shows up inside the active node. You can't "see" it from a global cluster member. I am now trying to add the listener but am hitting a dead-end.
# clrs create -Z test -g test-rg -t SUNW.oracle_listener -p ORACLE_HOME=/u01/app/oracle/product/10.2.0/db_snow -p LISTENER_NAME=test-ora-lsnr test-oracle_listener
clrs: taloraan.admin.uncw.edu:test - ORACLE_HOME /u01/app/oracle/product/10.2.0/db_snow does not exist
clrs: (C189917) VALIDATE on resource test-oracle_listener, resource group test-rg, exited with non-zero exit status.
clrs: (C720144) Validation of resource test-oracle_listener in resource group test-rg on node taloraan failed.
clrs: (C891200) Failed to create resource "test:test-oracle_listener".Is the LISTENER_NAME something that is assigned to the listener by Oracle or is it simply something of my choosing?
Also, how can I see a detailed status listing of the zone cluster? When I execute "cluster status", it doesn't give a verbose listing of my zone cluster. I have tried using "clzc status test" but am not afforded much in the way of output. As in, I would like to see some output that lists all of my resources.

Problems with Var. Bucket Consumpt. field in PPMs

Hi experts,
I'm trying to manually create a PPM (Production Process Model) into APO (to use it with SNP), however, the value entered in the
Bucket Consumption (Variable) field disappears when I active the PPM. Someone could tell me what I am doing wrong?
First, I created a Resource ( using fields of tab 1 Bucket ) with the following data:
Resource
Resource: R
Category (Cat): P - Production
Location: L
Time Zone: BRAZIL
Bucket Dimensn: Mass
Factory Calendar: BR
Bucket Capacity: 96000
Unit: KT - Kilotonne
Number of Periods: 365
Period Type: Day
Days -: 1095
Days +: 1095
ie, a resource with a capacity of 96000 Kilotonne per year.
The PPM I'm trying to create is the following:
PPM's General Data
Use of a Plan: S - PPM for Supply Network Planning (SNP)
Variable (Single Level Costs): 980,00
PPM's Operations and Activities
Operation: 0010:Produce
Activity: 10:Produce
Components of Activity
Product          I/O Indicator     Consump. Type     From Date     To Date          Unit     Var. Consumptn
A          O          E          01.01.1970     31.12.9999     KT     1000
B          I          S          01.01.1970     31.12.9999     KT     560
C          I          S          01.01.1970     31.12.9999     KT     180
D          I          S          01.01.1970     31.12.9999     KT     260
Mode of Activity
Mode     Primary Resrce          Location     Unit     Fix. dur.
10     R               L          D     4
Resource Mode
Resource     Location     Unit     Var. Bucket Consumpt.
R          L          KT     1000 (here is the problem)
PPM's Model
Prod. Process Model     Output Product     Date From     Date To          Planning Locatn     Discretization     Maximum Lot Size
PPM_1               A          01.01.1970     31.12.9999     L          x          9999999
ie, a PPM to produce 1000 KT of A in 4 days that consumes 1000 KT of R resource capacity.
The value of 1000 KT in the field Var. Bucket Consumpt. is not allowed. I could only register values between 0 and 9.
Thanks in advance for help.
Greetings,
Francisco Fonseca.
Edited by: Francisco Fonseca on Jun 26, 2009 1:56 PM
Edited by: Francisco Fonseca on Jun 26, 2009 1:59 PM

Hi,
Its a good practice that when you make PPM manually keep fixed duration as '1 Day' (in your mode of activity).
One thing i noticed that you made your resource as bucket resource, which is correct,
but the number you mentioned there (Bucket Capacity: 96000) is not right. What is the unit? It should be in hours (should be 8 or 9 or ).
That means in One day the resource is capable of producing goods for 8/9/ Hours. So, make the change there.
Now in your Mode of activity make fix duration 1 Day.
Components of Activity
Product          I/O Indicator     Consump. Type     From Date     To Date          Unit     Var. Consumptn
A          O          E          01.01.1970     31.12.9999     KT     1000
B          I          S          01.01.1970     31.12.9999     KT     560
C          I          S          01.01.1970     31.12.9999     KT     180
D          I          S          01.01.1970     31.12.9999     KT     260
Mode of Activity
Mode     Primary Resrce          Location     Unit     Fix. dur.
10     R               L          D     1
Now in resource: put following value
Resource Mode
Resource     Location     Unit     Var. Bucket Consumpt.
R     L     H          1
You will get an output as 8000 or 9000 (depending on your resource capacity per day, ie 8hr or 9 hr )
and if you put Variable consumption as 8/9 hrs the you will get 1000 KT per day, in that case you put variable material/product consumtion as 250 KT,
so you can get 1000 KT per four day. Variable resource consumption responsible for variable output (variable consumption of FG).
Pls revert if still there is any issue.
Thanks,
Satyajit
Edited by: Satyajit Patra on Jul 2
Edited by: Satyajit Patra on Jul 2, 2009 11:14 AM

Resource consumption in zones

Similar Messages

Maybe you are looking for