Workload Stats vs NoWorkload stats
What do you guys prefer? I understand noworkload is the new default but from reading about these, to me workload seems better no? Workload stats gather is against actual workload and noworkload is based on random reads and stuff.
Thoughts?
Gunners wrote:
What do you guys prefer? I understand noworkload is the new default but from reading about these, to me workload seems better no? Workload stats gather is against actual workload and noworkload is based on random reads and stuff.
Thoughts?Gunners,
NOWORKLOAD statistics are the default until the DBA collects system/CPU statistics. The system/CPU statistics only need to be collected once when the server is under a typical to heavy load (or after a change in the server's configuration). Take a look at the following from a test server (I posted this in another recent thread):
The following output shows a case where system statistics have not been gathered:
PNAME PVAL1 PVAL2
STATUS COMPLETED
DSTART 07-11-2009 00:04
DSTOP 07-11-2009 07:43
FLAGS 0
CPUSPEEDNW 2551.04281
IOSEEKTIM 10
IOTFRSPEED 4096
SREADTIM
MREADTIM
CPUSPEED 2752
MBRC
MAXTHR
SLAVETHR
The following shows a case where the system statistics have been gathered, but possibly with not a heavy workload:
PNAME PVAL1 PVAL2
STATUS COMPLETED
DSTART 08-01-2009 23:29
DSTOP 08-01-2009 23:29
FLAGS 1
CPUSPEEDNW 2545.45455
IOSEEKTIM 10
IOTFRSPEED 4096
SREADTIM .427
MREADTIM 3.782
CPUSPEED 2664
MBRC 32
MAXTHR 19181568
SLAVETHRAs you can see from the above, much more information about the server's performance is available to the cost-based optimizer when workload system/CPU statistics are available. Ideally, the optimizer should have as much information as possible to make good decisions when developing plans for queries.
You might want to take a look at the following articles to see how the system/CPU statistics are used by the cost-based optimizer:
http://richardfoote.wordpress.com/2009/11/25/the-cbo-cpu-costing-model-indexes-vs-full-table-scans/
http://jonathanlewis.wordpress.com/2007/04/30/system-statistics/
http://jonathanlewis.wordpress.com/2007/05/20/system-stats-strategy/
http://jonathanlewis.wordpress.com/2007/10/17/system-statistics-3/
Charles Hooper
Co-author of "Expert Oracle Practices: Oracle Database Administration from the Oak Table"
IT Manager/Oracle DBA
K&M Machine-Fabricating, Inc.
Similar Messages
-
System Stats opinions Please - WORKLOAD vs. NOWORKLOAD?
Dear Experts,
Bit of a background - We've a big (10TB) 10.2.0.5, 5 node RAC system for a big Oracle eBusiness Suite (R12 - HRMS/Payroll) implementation and at the minute we are running on some old (2 years old) gathered NOWORKLOAD system stats and since then a significant underlying IO infrastructure change has gone in and we are planning to upgrade the underlying server infrastructure from IBM P5 to IBM P7 soon as well (which introduces significant CPU changes etc.) and me personally think its the time to let Oracle know about all these changes under the carpet by updating the System Stats and since its likely going to be a significant change as potentially all the plans will be impacted (and few/some might go in an unintended direction) so want to know your opinion on these questions about it please :
1. Shall we just update the Noworkload Stats and leave it as it is (db_file_multiblock_read_count in left unset i.e. left for the system to set automatically).
OR
2. Shall we go for the WORKLOAD stats instead? If we go this way then I am planning to go by the brilliant strategy mentioned by Christian Antognini in his lovely book 'Troubleshooting Oracle Performance' for stability point of view.
3. We are planning to go 11g (11.2.0.2+) in the 2nd half to next year as well - is better to leave this evil till then if 11g introduces or improves the system stats mechanism significantly? An opinion favoring this is also that we'll get better and more robust testings done if we pile it on 11g project so may be a bit safer approach (and better for work/human resources planning as well).
Out of the 5 RAC nodes, 2 are dedicated to OLTP stuff, 2 for Batch (Payrolls etc.) and 1 for (low priority) Reporting so obviously there are concerns around getting to a setting that's going to be the best fit for all of them and since thats not easy to achieve the risk is higher.
I am sure I am not new here and some of you'd have come across these questions as well so am wondering if I can get some opinions or advices please? Particularly from the people who have gone from NOWORKLOAD to WORKLOAD (preferable eBusiness suite and even better HRMS/Payroll) please and also from the people who've distributed workload like we have on difference RAC nodes- if they can share their experiences on how did it go for them - that'll be good please? (Obviously everyone's system is different and so is everyone's workload but if 8 out 10 say its been good for them or even bad - then that gives a bit of a feel about it).
Or are most people still running on 'safer' NOWROKLAOD stats please becs WORKLOAD is deemed to be a bigger change?
Also, in the end - I've heard some not so great stuff about IBM P7 as well - just wondering if there are people out there with some experience on these as well please?
Thanks in advance.pdthedba wrote:
3. We are planning to go 11g (11.2.0.2+) in the 2nd half to next year as well - is better to leave this evil till then if 11g introduces or improves the system stats mechanism significantly? An opinion favoring this is also that we'll get better and more robust testings done if we pile it on 11g project so may be a bit safer approach (and better for work/human resources planning as well).Be careful about gathering system statistics in 11.2.0.2. Why? Well there is a bug.
Workload System Statistics Bug in 11.2
However, as of earlier this year there is a patch to fix it: 9842771. -
How to change train step from visited state to disabled state
Hi all,
I like to change visited train step into disable state again.
Is there anyway to do this?
With Regards,
WPHi M.Jabr,
Thanks for your reply.
It works.
i like to know one more thing about train.
is there anyway to change visited state to unvisited state?
With Regards,
WP -
Converting an update statement to modify statement in BADI
Hai I want to convert an update statement to nodify statement . Can we define the internal table in BADI. Give me the code
Hi,
We can define internal tables in BADI.
Example:
Implement BADI ME_PROCESS_PO_CUST and in the method PROCESS_ITEM
DATA: ls_mepoitem TYPE mepoitem.
*get item data
ls_mepoitem = im_item->get_data( ).
IF ls_mepoitem-adrn2 EQ space AND
ls_mepoitem-adrnr EQ space.
Retrieve the adddres or hardcode as per ur reqmt.
*transfer the address to ls_mepoitem-adrn2.
MOVE v_adrn2 TO ls_mepoitem-adrn2.
CALL METHOD im_item->set_data
EXPORTING
im_data = ls_mepoitem.
ENDIF.
Asha -
Statement and Prepared Statement
if i write a same query with statement and preparedStatement
like
Select * from emp_detail where emp_id =20;
stmt= con.createStatment();
rs = stmt.executeQuery(query);
and using preparedStatement
Select * from emp_detail where emp_id =?;
pstmt= con.prepareStatement();
pstmt.setInt(1,20);
rs = stmt.executeQuery(query);
Using which statment(Statement or Prepared Statement) the data will retrive fast and why.... in these case ????Statement should be quicker than Prepared Statement, because
Statment do only one thing: send that sql to server or run that sql directly.
Prepared Statement should do two (or more than two)things:
1. parse your sql first, prepare a store procedure, then call that store procedure.
Or
2. prase your sql first, then use your value to replace "?" for getting a new sql, then work like Statement.
Prepared Statement is quiker when you use it repeatedly. -
Statement,Prepared Statement and callable statement
Hi,
Please let me know in which scenario we are using Statement,Prepared Statement and callable statement.
and which is efficient one among the above.
Thanks in advanceWelcome to the forum!
>
Please let me know in which scenario we are using Statement,Prepared Statement and callable statement.
>
We don't know what scenario you are using those in or if you are using them at all. Are you asking what they are?
For document related questions you should consult the documentation or use your favorite search engine to get information.
See the Java Tutorial
http://docs.oracle.com/javase/tutorial/jdbc/basics/prepared.html
>
The main feature of a PreparedStatement object is that, unlike a Statement object, it is given a SQL statement when it is created. The advantage to this is that in most cases, this SQL statement is sent to the DBMS right away, where it is compiled. As a result, the PreparedStatement object contains not just a SQL statement, but a SQL statement that has been precompiled. This means that when the PreparedStatement is executed, the DBMS can just run the PreparedStatement SQL statement without having to compile it first.
>
The Javadocs for your Java SDK have the API for each of those classes and a description of what they are. And the Oracle JDBC Developer Guide has extensive information on how to use them.
http://docs.oracle.com/cd/B28359_01/java.111/b31224/toc.htm -
Cash Flow Statement & Funds Flow Statement
Hi,
Good day!
I'm currently working on a project where I have to generate a Cash Flow Statement & Funds Flow Statement from SAP. I would just like to ask how to generate the same and what tables or databases I need to consider or check in preparing it.
Regards,
Nimish AgarwalDear Colleagues:
I would like to ask you if you know a Report that cover the requirement of "State of origin and application of Funds" Report. I would need an Standard Report in SAP in order not to make a Z development.
Your Sincerely
leonardo lopez
FI Consultant. -
Retained Earnings Statement Format(Financial Statement Version)
Dear Friends,
Since 2 days i am requesting you regarding Retained Earnings Statement Format(Financial Statement Version)
Please send me the fromat for Retained Earnings Statement Format(Financial Statement Version)
It would be a great help for me and points will be assigned...
Regards
Sap Guru
[email protected]hi sapguru,
please cheek your mail and assign points. -
How to create active state or current state in spry menu ?
i have build a vertical menu with spry menu widget, Just
wondering if any one can guide me how to create an active state or
current state in the menu.
I want the visitor to understand which page they are on, so
just want to change the colour of the active link.<helpful but self-serving too>
There are several ways to do this, but the easiest way, if I
may say so,
is to use my divaGPS Extension. It works with virtually any
menu type
and is designed to accomplish precisely what you are asking
for quickly
and easily.
</back to our regularly scheduled programming>
E. Michael Brandt
www.divaHTML.com
divaPOP : standards-compliant popup windows
divaGPS : you-are-here menu highlighting
divaFAQ : FAQ pages with pizazz
www.valleywebdesigns.com
JustSo PictureWindow
JustSo PhotoAlbum -
How to highlight State of Multi-State object?
hi,
I have created 3 states of object State1, State2, State 3. and want to highlight state in object state panel programmatically. can any one have idea abt it?@Priyanka – you did not mention a specific scripting language.
In ExtendScript (JavaScript) you can set the "active" property of the 3rd state of your MultiStateObject to value "true".
//ExtendScript (JavaScript) example
//Make sure a MultiStateObject is selected; nothing else, then run this JavaScript:
app.selection[0].states[2].active = true;
The expression from left to wright (for lurkers who never saw a line of code):
app => application (that's InDesign, of course)
selection => your current selection in the UI (could be one item, could be many)
[0] => index number 0 is the first item in your selection; be cautious with the expression "first", it depends on the InDesign version what that does exactly mean, if you selected more than one object
states => obviously the "state" object (always with the plural "s", it's the collection of all states in that selected MultiStateObject)
[2] => index number 2 is the 3rd element of the states collection
active => property for the active status; boolean, could be "true" or "false"
= => with the "=" sign, you are setting the value
true => the value of the property; self-explanatory, I hope…
; => end of expression, all that comes after is a different expression.
The different parts are held together with a dot-notation. The index is directly glued to its object, no dot notation between object and index.
If you want learn more about ExtendScript and the InDesign DOM see Jongwares' documentation at:
http://www.jongware.com/idjshelp.html
Uwe -
Data between states of the states machine
So I have two tables that I generate in one state of the state machine. My question is what is a proper, aesthetic way to propogate these two tables into one another state of the state machine. Please see the attached code. State machines rock!
Attachments:
flag.jpg 166 KBThe shift register is the right tool for this job. It follows the left to right data flow paradigm, unlike other solutions (such as using a local variable). I am sure you will get a few other opinions but I think that the shift register is the way to go.
CLA, CLED, CTD,CPI, LabVIEW Champion
Platinum Alliance Partner
Senior Engineer
Using LV 2013, 2012
Don't forget Kudos for Good Answers, and Mark a solution if your problem is solved. -
Bdoc state - Received (intermediate state)
Hi experts,
Can you please help me with this. We've just found out that we have several bdocs with state "Received (intermediate state)" in middleware. They said that if these are in this state for quite some time, we need to reprocess it to prevent any future problems. When I choose one and click on Show Bdoc Msg Errors/Receivers, it only says "No receivers or error segments for Bdoc message found."
Can you please let us know how to clear these? I'm providing below two bdocs on these state that may help you understand the problem.
1.
Bdoc state description: Received (intermediate state)
Bdoc type: BUPA_REL
Flow context description: mBdoc Validate
Send date: 01.09.2010
Send time: 19:01:06
Sender site name: (This is our ECC box)
Queue name: R3AD_CUSTOMExxxxxxxxxx (x > customer number)
User (Creator): MWRFCXXX (x > our ECC box)
2.
Bdoc state description: Received (intermediate state)
Bdoc type: CND_M_SUP
Flow context description: mBdoc Validate
Send date: 09.02.2011
Send time: 15:43:04
Sender site name: (This is our ECC box)
Queue name: R3AD_DNL_COND_A594
User (Creator): MWRFCXXX (x > our ECC box)
Hope you can help us here. Thanks a lot!
Kind regards,
MikeHi,
Go to transaction R3AR3,Are all your requests still there.IS there any queue stuck in smq2 and SMQ1.If there is nothing stuck in SMQ1 or SMQ2,then i think for you Bdoc type BUPA_REL,for all those queues in intermediate state there is no number found for the range which you used to create request for Partner relationship.
Regards,
Arya -
Hi..
I am trying to decode some values using java.nio api but throwing the following exception ...
Can any one explain why that exception comes and what are that states mentioned in exception...
Exception :
Caused by: java.lang.IllegalStateException: Current state = FLUSHED, new state = CODING_END
at java.nio.charset.CharsetDecoder.throwIllegalStateException(Unknown Source)
at java.nio.charset.CharsetDecoder.decode(Unknown Source)
Code :
CharBuffer charBuffer = CharBuffer.allocate( 1000 );
charBuffer.clear();
CoderResult coderResult = charsetDecoder.decode( input_byteBuffer, output_CharBuffer, true );
workCharBuffer.flip();Ditto. The pattern I get can be summarised as:
15 IllegalStateExceptions per second for about 2 minutes, then:
"ContextManager: Removing context Ctx( /mycontext )", then:
java.lang.NullPointerException
at org.apache.tomcat.core.ContextManager.handleError(ContextManager.java:1177)
at org.apache.tomcat.core.Handler.service(Handler.java:312)
at org.apache.tomcat.core.ServletWrapper.service(ServletWrapper.java:372)
at org.apache.tomcat.core.ContextManager.handleError(ContextManager.java:1177)
at org.apache.tomcat.core.Handler.service(Handler.java:312)
at org.apache.tomcat.core.ServletWrapper.service(ServletWrapper.java:372)
at org.apache.tomcat.core.ContextManager.handleError(ContextManager.java:1177)
(Infinite loop repeats until Java poos itself)
System: Apache 1.3.26, Tomcat 3.3 and JDK 1.4.1_02
The whole setup works JUST FINE using JDK 1.3.1_07
Any ideas? I can't upgrade to Tomcat 4.1, since the overhead of meeting Servlet 2.3 makes it run like a dog, compared with 3.3, for a non-JSP site. A whole other story . . .
Cheers, Chris -
Number of actual state of multi-state objects
Hello,
is it possible create automatic number of actual state and automatic total number of states in mulsti-state objects?
Thank you for your answer.
Martin@Martin – the following script will show you an alert message of the active state of all first-level MSOs in the active document (it will not show nested MSOs like in groups, MSOs anchored to text frames, MSOs inside MSOs etc.pp.).
Screen shot of the result message, that can be selected and copied to the clipboard for further use:
Code is ExtendScript (JavaScript):
//ShowActiveStatesForAllFirstlevelMSOs_DOCUMENT.jsx
//Uwe Laubender
//See discussion at:
//http://forums.adobe.com/thread/1365865?tstart=0
* @@@BUILDINFO@@@ ShowActiveStatesForAllFirstlevelMSOs_DOCUMENT.jsx !Version! Sat Jan 04 2014 11:41:46 GMT+0100
if(app.documents.length === 0){alert("No document open. STOP.");exit()};
if(app.documents[0].multiStateObjects.length === 0){alert("No First Level MSOs found. STOP.");exit()};
var d=app.documents[0];
var allLevelOneMSOs = d.multiStateObjects;
var allLevelOneMSONames = d.multiStateObjects.everyItem().name;
var allActiveStates = d.multiStateObjects.everyItem().activeStateIndex;
var numberOfStates = new Array();
var nameOfActiveState = new Array();
for(var n=0;n<allLevelOneMSOs.length;n++){
numberOfStates[n] = allLevelOneMSOs[n].states.length;
nameOfActiveState[n] = allLevelOneMSOs[n].states[allLevelOneMSOs[n].activeStateIndex].name;
var resultArray = new Array();
for(var n=0;n<allLevelOneMSOs.length;n++){
resultArray[n] = [
"\""+allLevelOneMSONames[n]+"\""+
"\t"+"\""+nameOfActiveState[n]+"\""+
"\t"+"is active state: "+"\t"+(allActiveStates[n]+1)+" of "+numberOfStates[n]
resultArray;
scrollableAlert("Active states of all First Level MSOs in document",resultArray,false,"Cancel");
function scrollableAlert(titleString/*String*/,inputArray/*Array*/,listActive/*[true] || [false]*/,buttonText/*String*/){
//Basic function by Peter Kahrel, enhanced by Marijan Tompa [Tomaxxi] at: http://tomaxxi.com/2010/07/scrollable-alert/
//Further enhanced by Uwe Laubender with a additional argument: listActive, buttonText
var inputArrayAsString, w, list
if(inputArray instanceof Array){
inputArrayAsString = inputArray.join ("\r")
else{inputArrayAsString = inputArray};
w = new Window("dialog", titleString);
list = w.add("edittext", undefined, inputArrayAsString, {multiline: true, scrolling: true});
list.maximumSize.height = w.maximumSize.height-100;
list.minimumSize.width = 540;
list.active = listActive;
w.add("button", undefined, buttonText, {name: "ok"});
w.show();
}; //END OF function scrollableAlert()
If you have questions about this script or want a script that is more functional, just send me a personal message (PM) via this forum.
Uwe -
Difference between "print" statements and "select" statements , TSQL
What is the difference between "print" statements and "select" statements when it omces to debugging and watching varibles in the stored procs .....
SELECT statement is part of the ANSI SQL language.
PRINT command is not part of the SQL language, it is used for debugging in T-SQL.
BOL:" Returns a user-defined message to the client.
A message string can be up to 8,000 characters long if it is a non-Unicode string, and 4,000 characters long if it is a Unicode string. Longer strings are truncated. The
varchar(max) and nvarchar(max) data types are truncated to data types that are no larger than
varchar(8000) and nvarchar(4000).
RAISERROR can also be used to return messages. RAISERROR has these advantages over PRINT:
RAISERROR supports substituting arguments into an error message string using a mechanism modeled on the printf function of the C language standard library.
RAISERROR can specify a unique error number, a severity, and a state code in addition to the text message.
RAISERROR can be used to return user-defined messages created using the sp_addmessage system stored procedure."
LINK: http://technet.microsoft.com/en-us/library/ms176047.aspx
The SSMS (client software) returns SELECT output to Results and PRINT output to Messages.
The SQLCMD -o (output file) option captures all outputs into one file mixed.
Kalman Toth Database & OLAP Architect
T-SQL Scripts at sqlusa.com
New Book / Kindle: Exam 70-461 Bootcamp: Querying Microsoft SQL Server 2012
Maybe you are looking for
-
Where did my messages go? (Only "showing" 50 messages).
I have had my iPhone 3G set to show 50 Recent Messages (e-mail). I have had about 20 messages in my InBox, but then had more than 30 new messages this morning. Some of the older messages in my InBox have disappeared. I checked in the Trash, but they
-
Hi ! I am creating a report in Reports 3.0. I have to print some 25 fields which will wrap up in 2/3 lines. My problem is that I am not able to increase the output width from 80 to 132. My output should go to a file in the hard disk of my computer. F
-
Customize colors in horizontal bar in charts OBIEE 10g
Hi All, I have a chart report which has horizontal bars. I want th change the color of each bar to different colors. I have no condition,any color is fine for me. How can I achieve it? In Format Chart Data, I haev specified different colors,but it gi
-
[SOLVED] vfs_zip plugin for deadbeef doesn't work
hello everybody here, i found vfs_zip plugin here doesn't work, that is, i can import music files from zip archive, but i can't play them- if i play those ones, the deadbeef'll be stuck ,even if i try play other files. so, any solution here? any idea
-
Why am i redirected to the page that says "set up icloud on this device" when i click the icloud sign in ? I am positive im signed up