Execution speed question

Does the order I type criteria in the WHERE clause have any bearing on how fast the query will execute? Or, does Oracle ignore this completely and try to optimize the query regardless of the order I put the criteria in?

Here's a pertinent section from that document I linked you to....
"The ORDERED_PREDICATES hint forces the optimizer to preserve the order of predicate evaluation, except for predicates used as index keys. Use this hint in the WHERE clause of SELECT statements.
If you do not use the ORDERED_PREDICATES hint, then Oracle evaluates all predicates in the following order:
1. Predicates without user-defined functions, type methods, or subqueries are evaluated first, in the order specified in the WHERE clause.
2. Predicates with user-defined functions and type methods that have user-computed costs are evaluated next, in increasing order of their cost.
3. Predicates with user-defined functions and type methods without user-computed costs are evaluated next, in the order specified in the WHERE clause.
4. Predicates not specified in the WHERE clause (for example, predicates transitively generated by the optimizer) are evaluated next.
5. Predicates with subqueries are evaluated last, in the order specified in the WHERE clause."

Similar Messages

Why is the difference in execution speed of the function "SetCtrlVal" between constant and changing values so small ?

In my large application (1 MB exe-file) I am continuously updating a lot of numeric controls with new values. Most of them do not really change their value. Within my search of improving the performance of my application I noticed, that there is only a small difference of the execution speed between a call of "SetCtrlValue" with constant values and calls with changing values. It runs much faster (25 times on my PC), if I get the actual control value with "GetCtrlVal", compare it with my new value an do a call to "SetCtrlVal" only if the current value and the new value are different.
My questions to CVI-developers is:
Isn't it possib
le to do this compare within the function "SetCtrlVal"
My question to all CVI-users is:
Does anyone have similar tips to improve the performance of CVI applications ?
I developed a small test application for this problem, which I can mail to interested users.

What takes the extra time is the redraw of the control. When you call SetCtrlVal we ALWAYS redraw the control. We wouldn't want to build in functionality to check if the value was the same because that would add additional time to the SetCtrlVal in every case. If you want to do it outside of the loop you can as you have done above. You have a few options. First, keep a previous value variable for the controls that you can use to determine whether to set the control value. I.E.
int oldVal = 0;
int newVal = 0;
if(newVal!=oldVal) {
SetCtrlVal(..., newVal);
oldVal = newVal;
Also, if you set the value of a control through SetCtrlAttribute instead, there is no built in redraw of the control (which is what takes all the time). Using SetCtrlAttribute
to set the value is very fast, but remember there isn't a built in redraw on the screen to display the new number.
Best Regards,
Chris Matthews
Measurement Studio Support Manager

Execution speed improvements of FGV over locals

This my first post. I am new to LabVIEW and am currently writing my first significant application. I am actually modifying an existing application, but am adding additional functionality.
I have read this VI Execution Speed article but still have questions http://zone.ni.com/reference/en-XX/help/371361H-01/lvconcepts/vi_execution_speed/
As an example, I am populating an xy plot with several sets of data, such as saved data points and a curve fit. I am currently using a case structure to decide when to write various elements. As an example, I only write the curve fit one time and it remains static as other pieces of data are added to the graph. In an attempt to make things faster I did not want to redo and redraw the fit each time. I am using locals within the case to populate the graph. In the default case nothing if written to these locals (and I assume the graph is not redrawn)?
I now realize that instead using a case in which nothing is written, I could use a feedback node (to write the previous case). The in addition, in place of using a local, I could use a Functional Global Variable.
Would it be better to have several cases which write locals to populate the plot and one case which writes nothing OR Have the several cases write to a FGV and then a default case which writes the last data to the FGV via a feedback node?
The first seems like less load as nothing is written in the default case, but the other cases do write to a local which has a front panel object that I don't need or want.
Iis there any benefit to hiding unused front panel objects?
Is there a way to create a local without front panel object?
Finally, with a tabbed UI, are objects on the non active tabs being redrawn and slowing the process?

Good questions!
Starting from the bottom:
Iis there any benefit to hiding unused front panel objects?
Is there a way to create a local without front panel object?
No and No. Local variables are generally a poor way to store and move data. For the reasons you cited (FP object you do not want or need, hidden FP objects) and that they force a copy of the data and may cause a thread switch to the UI thread.
Finally, with a tabbed UI, are objects on the non active tabs being redrawn and slowing the process?
It depends. Newer versions of LV are generally smart enough that non-visible indicators are not redrawn. This may not have been true in older versions, although I do not know when the changes were made.
You certainly do not need to redo the fit if the data has not changed. I think the entire graph gets redrawn when any new data is written.
Search for Ben's extremely informative Nugget on Action Engines (AE). An Action Engine is a FGV with added capabilities. It may be a very good option for what you are trying to do. It could store all the data sets, update the graphs, and make the data available where it may be needed without extra copies. With the AE you could easily acquire and store data at one rate and update the graphs at a (slower) rate appropriate to the users eyes and brains. Updating graphs more than ~10 times per second is a waste of resources because the user cannot respond any faster than that. Also you could write a subset or reduced data set to the graph if the amount of data is larger than the number of pixels in the plot.
Lynn

Execution Speed help

Hi there!
I hope someone can help me about my (probably noobish) question:
I have 2 Oracle environments, both not maintained by me and if I execute the following querry:
select * from DATA_TBL where rownum < 11;
It takes about 8 seconds on the one and less than 1 second on the other environment.
As there is everything selected via '*' and the where clause does not include data colums it can't be about indexes or analysed information or?
On both environments I also get the same explain_plan (only thing I have the rights to do):
PLAN_TABLE_OUTPUT
| Id | Operation | Name | Rows | Bytes | Cost |
| 0 | SELECT STATEMENT | | | | |
|* 1 | COUNT STOPKEY | | | | |
| 2 | TABLE ACCESS FULL | DATA_TBL | | | |
Predicate Information (identified by operation id):
1 - filter(ROWNUM<11)
Note: rule based optimization
15 rows selected.
I would be happy if someone could tell me about possible reasons for the difference in execution speed, or what to tell the DBA's..
Thanks in advance!
Jan

>
select * from DATA_TBL where rownum < 11;
It takes about 8 seconds on the one and less than 1 second on the other environment.
>
SQL> create table t1 as select 'AAAAAAAAAAAAA' as word from dual connect by level<=1e7;
Table created.
SQL> create table t2 as select 'AAAAAAAAAAAAA' as word from dual connect by level<=1e7;
Table created.
SQL> delete from t1 where rownum<5000000;
SQL> commit;
Commit complete.
SQL> set timing on
SQL> select * from t1 where rownum<10;
WORD
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
9 rows selected.
Elapsed: 00:00:02.44
SQL> select * from t2 where rownum<10;
WORD
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
AAAAAAAAAAAAA
9 rows selected.
Elapsed: 00:00:00.48Kind regards
Uwe
http://uhesse.wordpress.com

Execution speed onboard programs

Hi all
I am tying store the value of analog input channels of PCI-7352 on buffer, from onboard program.
The program run fine, but I don´t know how measure the total acquisition time in the program.
In an example in LabView ( adc-gpbuffer.vi ), use 'flex_load_delay' function to establish an interval in
the acquisition. But the real interval between readings not is calculated.
Another question, How I can know the execution speed of onboard program ?. The documentation does not
show information.
Thank´s
Javier
Attachments:
adc-gpbuffer.vi ‏165 KB

Javier,
there is no easy way to measure the execution timing of onboard programs. One thing that you could do is to toggle a digital line in your onboard program and measure the timing with an oscilloscope.
Please be aware that the timing of onboard programs doesn't work deterministicallly as onboard programs don't run with time critical priority on the board's CPU so you will probably see a fair amount of jitter.
The main purpose of the analog inputs on the 7352 is analog feedback. You can use them for single point measurements, too but if you need to acquire data with an accurate timing you better should use an additional M-Series board like the PCI-6220 which provide a much better measurement and timing accuracy and as true measurement devices they provide a whole set of additional useful features.
Best regards,
Jochen Klier
National Instruments Germany

EXECUTION SPEED OF FINISHED PROGRAM

Thanks Guys, You comments are really encouraging.
I agree Java is great!
My question really is put better this way:
The Execution speed of finished Java programs is slow and often does require much memory? Why is this and what are the solutions?
Thanks again

The Execution speed of finished Java programs is slow and often does require much memory? Why is
this and what are the solutions?Well, the solution is simple then: leave those Java programs unfinished; classic composers did the
same and they got famous for it ... Seriously though, your question is too vague; what is 'slow' and
what is 'much memory'? Before the application was 'finished' did you do any (unit) tests? It can't be
as if suddenly you go "my golly, it's slow and big now it's finished; lets complain". Care to elaborate?
kind regards,
Jos (fast, small and so soft for your hands ;-)

How do I associate timing with multiple AI loops that is independant of loop execution speed

I am using Labview 7.1 and I am performing AI of
voltage on different channels of the same multifunction DAQ. I am
using while loops to aquire the data. I am writing the data to
spreadsheet file with an associated time as a 2D array. Currently I am
using the 'elapsed time' interactive subVI to get the time in seconds
(attempting to use the 'Get Date/Time in seconds' and converting it to
DBL results in time values that do not change. My thoughts are that
maybe the number of seconds is too large to display single second
precision with a DBL floating point number). When I run AI on two
multiple loops they execute at different rates and thus the time values
from "elapsed time" are accumulated at different rates. I have tried
using timed loops to control timing, but if one loop executes at a rate
slower than the timing of the loop the 'elapsed time' still accumulates
at different rates. I need to be able to associate both analog inputs
to the same time
value in the spreadsheet value and I would like to find a way to
associate a time with each data point that is independant of loop
execution speed (although I would still like to control execution speed
of the loops. I am pretty new to Labview and programming in general,
any help would be greatly appreciated.

If you want to read more than one AI at the same time, you should use a trigger. You would need to set up the AI Trigger to an external source. Then you would have to use a function generator or digital out to create a clock. The clock would be wired to the AI external trigger. When the clock goes high (or low depending on configuration), both AIs would read. That is the method I use when needing to sync multiple AI inputs.
- tbob
Inventor of the WORM Global

Execution Speed of Aurora JRE

We are experiencing some problems with the execution speed of the aurora JRE. The documentation states that the Aurora JRE is compiled to native code and should run 2 - 10 times faster than a normal JRE. We are atempting to do some memory sorts using the Array class which is taking 10 times longer in the Oracle instance than on a client.
Is this a known performance bottleneck, are there any "tweaks" available?
Thanks
Julian

Hello Ravi,
Does your statement about core java classes being natively compiled also apply to the JVM in OAS / iAS?
I remember reading in a white paper that, in iAS, java code would be translated into C code for native execution. Does this apply to user code or only core java classes ?
(by the way, It seems impossible to download OAS 4.0.8.2 from OTN, maybe it would help if it was divided into several smaller files...)
Thanks, Remi DEH
<BLOCKQUOTE><font size="1" face="Verdana, Arial">quote:</font><HR>Originally posted by Oracle Support Analyst (Ravi):
In the current JServer release, Java code you load to the server is
interpreted. The underlying core classes upon which your code relies (java.lang.*) are natively compiled. Until the native compiler is available for user programs,
the net speed benefit of native compilation to your executing program is dependent upon how much native code is traversed, as opposed to interpreted code. The
more Java code from core classes and Oracle-provided class libraries you use, the more benefit you will see from native compilation.
In 8.1.7 i.e 8i Release 3 we support natively compiled code for the user programs.
<HR></BLOCKQUOTE>
null

Execution Speed of my page

Hi everybody,
I converted an existing report into an FM and used that FM in my BSP page to display the output in browser.
But the execution speed of BSP page is almost half as compared to the execution of report. I am not sure why this has happened.
Is the layout reponssible for this low speed?
But its excatly the same layout that report displayed.
Please tell me where should i do the changes to increase execution speed .
Thanks

I suggest you to execute SE30 transaction (performance measurements) and including your BSP. You can do this also from SICF, I think that in the menu you hace a Runtime Analysis option or similar (sorry I by now I cannot acces to my SAP BW). Then, check the report, and discover where the most of the time is spent.
I had a similar problem. I had a BSP which execute many times a FM for generate a data table. I put AJAX to my BSP to improve performance and concurrency.

VI execution speed depending on mouse movement

Hi,
I have a large application that sometimes became very very slow. When It happens the only thing that I was able to notice is that moving mouse, execution speed of vi come back normal. I checked all the opened reference but I didn't find any not closed reference.
Any tip will be helpfull..
Golzio

can you give us any other information or post an example of the program. Does it use an event structure, data acquisition, are you performing file i/o? If you can give us an example it would be most helpful. Thanks
BJD1613
Lead Test Tools Development Engineer
Philips Respironics
Certified LV Architect / Instructor

VI execution speed in subpanel is slow

Hello,
I have the following problem with subpanels. I have a VI that should run inside of a subpanel. This VI contains an Image Display component (Vision). This Image Display is updated with a new image inside a while loop every 40 milliseconds. This is the one and only action inside of this loop. When I run this VI in a subpanel, the execution slows down to 200 ms for the Image Display update action (for the while loop). I measured the this with a Timervalue (ms) component. If my application runs the VI outside of the subpanel in an own window then the timig of 40 milliseconds works perfect. The only difference between the two situation is that the VI runs in or outside of a subpanel.
This is a great problem for me because my camera captures images with a framerate of 7,5 fps. So an Image Display update time of 200 ms is not acceptable for my application. Is this a known issue. Is there a workaround or some settings that I should tweak to improve execution speed of VIs in a subpanel. Or should i simply not use subpanels if execution speed matters - then it would be good to write this into the subpanel documentation. I carefully read the documentation an searched the dicussion forum but I could not find an answer.
Thank you very much for any help.

Hey Incredible,
could you tell me your labview version and please post your labview code.
Kind regards,
Elmar

Execution speed and Print Statements

I use so many "System.out.println" statements in my java program.
Will it reduce the execution speed of the application?

Everything that does something will make the application slower.
You should instead ask yourself if you need the print statements, and if the application is fast enough. You can also execute your application in a profiler to find bottle necks.
Kaj

Another question about query execution speed.

I know I should normally start with posting the query but my question will take a little different direction, so here is the deal:
We have an Apex application, there is a query on one of the pages that for certain users takes quite long, lets say 1min to 2 min, depending on how many records the query returns. The actual query varies depedning on some security settings and other parameters on the page - it's constructed by pl/sql function returning sql.
By querying v$sql I get the exact sql run by the user for who the query takes long time to execute.
Then I run that same sql statemt in Toad and it always takes 2 seconds to execute. For the user executing the same sql from Apex takes consistently 50-100 sec. My Toad is set to return up to 10,000 records and the query returns less than 500.
Can somebody help me make some sense of this?
George

Hi there
geo2 wrote:
We have an Apex application, there is a query on one of the pages that for certain users takes quite long, lets say 1min to 2 min, depending on how many records the query returns. The actual query varies depedning on some security settings and other parameters on the page - it's constructed by pl/sql function returning sql.To render a page, Apex does much some more stuff than just run the query. For instance, depending on the type of pagination you have enabled it needs to get the total number of rows to display "x to y of z" (or to display the message "more than nnn rows found"). Have you enabled debugging in Apex to be sure that most of the waiting time is spent on that query?
By querying v$sql I get the exact sql run by the user for who the query takes long time to execute.
Then I run that same sql statemt in Toad and it always takes 2 seconds to execute. For the user executing the same sql from Apex takes consistently 50-100 sec. My Toad is set to return up to 10,000 records and the query returns less than 500.Does Toad takes only 2 seconds to retrieve all rows or just to get the first top rows?
Luis

For r in (select) loop: execution plan question

Hi there,
I have a procedure which does the following (on Oracle 11.1.0.7):
begin
for r in (select col1,max(col2) as col2 from tab@remote_db group by col1 ) loop
    update tab2
       set col_a = r.co1
       where col_b = r.col2;
    commit;
end loop;
end;The subselect from the remote table delivers about 4000 rows.
When executing fast, the procedure runs about 20 seconds, the trace shows that the select is executed 1 time, fetches 49 times. The update executes 4000 times.
When executing slow, the procedure runs about 20 minutes (!), the trace shows that the select is executed 1 times, fetches 4000 times. The update executes 4000 times.
This happens without changing the code, from the same client (sqlplus on the server, same version as database).
Can anyone probably give me a clue how I could track down the plsql execution steps like the optimizer trace that let you see why a certain sqlplan was not taken or what might cause that change in behavior?
I cannot produce that behavior at will - most times the procedure runs fast.
thanks,
Heimo

Hi again,
regarding the "faulty test" and what is fastest: user tested both versions and came up with the for-loop to satisfy him best. I would, of course, utilize event 10046 and tkprof and a repeatable dataset to have solid data. Sorry for not setting enough emphasis on my real question, which is described in the following.
But, and that is what is really causing awkward feelings for me: when doing plain sql, I can use events 10046, 10053 and learn from the tracefiles why a decision was taken (now that is making me feel good, actually). But when it comes to plsql, I have not yet such a handy toolset that helps me understand what influences the plsql-engine to go this or that way.
Thus, I valued the posts pointing to dbms_trace and plsql_optimization_level as helpful answers.
Because of that I learned about the views %_PLSQL_OBJECT_SETTINGS and remembered to check for other plsql-initora settings.
I also verified that neither the 10046 nor the 10053 trace files contain any reference to plsql-parameters/settings, and that the 10053 trace only shows the subselect and update, but no info if it would "bulkify" or not.
Meanwhile I did start to utilize dbms_trace (which causes the statement to go the unbulkifyed by the way) and maybe I can learn from that output or from following posts?
best regards
Heimo

BT Infinity Noise/Speed questions

Is my noise margin good or bad?
1. Product name:
BT Home Hub
2. Serial number:
+068543+NQ42955553
3. Firmware version:
Software version 4.7.5.1.83.8.173.1.6 (Type A) Last updated Unknown
4. Board version:
BT Hub 5A
5. VDSL uptime:
0 days, 19:59:25
6. Data rate:
16747 / 65657
7. Maximum data rate:
20145 / 65379
8. Noise margin:
5.9 / 5.7
9. Line attenuation:
0.0 / 17.7
10. Signal attenuation:
0.0 / 17.4
11. Data sent/received:
376.0 MB / 2.4 GB
People say its suposed to be around 6db so why is my speed so low?
It used to be 77mb last week. Does the speed reduce by how many customers are on the cabinet? We just got the FTTC installed so lots of people are buying it just now.

If you use this checker it will tell what cabinet you are on and what services it can receive. post back here editing out your phone number
http://www.dslchecker.bt.com/adsl/adslchecker.welcome
If you want to say thanks for a helpful answer,please click on the Ratings star on the left-hand side If the reply answers your question then please mark as ’Mark as Accepted Solution’

Execution speed question

Similar Messages

Maybe you are looking for