Commit Block Question

What is the best method to determine the size of the commit block?

What is the best method to determine the size of the commit block?

Similar Messages

  • Commit blocks

    Hi!
    Don't know if this is a ORACLE feature or not but...
    Here is my problem:
    I have an application that receive information from a source, process
    the data and input the result in a data warehouse.
    The problem is that sometimes the commit blocks, i.e., when I'm making
    the commit the application stops at that point.
    I think that the problem as something to do with materialized views. In some of then I use "REFRESH ON COMMIT" and in other cases "REFRESH ON DEMAND".
    The version of Oracle is 9i.
    Thanks in advance,
    Carlos Dias

    The usual cause is when another transaction has locked the records you want to change. Make sure no others sessions are making transactions to the same table. Even when viewing data,some client tools lock the table, and the commit hangs.

  • Calc optimization issue with commit blocks

    Hi there,
    How can we know the right commit block interval setting to the database?
    and though i am doing parallel calc its forcing to do the calc in serial mode due to some dependencies.
    Can anyone suggest me the best way to remove the dependencies!!!
    I am having lot of member formulas in my calc so its forcing the topdown cal mode though i am doing bottumup....
    How to avoid all these dependencies...or do u think topdown should be fine?

    It is better to use the default setting in the transactions tab.
    you can understand the behaviour of thses settings in the below link
    http://download.oracle.com/docs/cd/E10530_01/doc/epm.931/html_eashelp/frameset.htm?dbdprptr.htm
    Mostly dense formula calculations will have calc dependencies and force the serial calculations.
    Tag all the dense formula members dynamic calc wherever possible.
    If any thing needs to be calculated in the batch, calculate by fixing Lev0 sparse combinations.
    In the Database rollup, Use Calcdim of sparse dimensions only.

  • How to type comma "," and question "?" without going to "123" sub menu?

    How to type comma "," and question "?" without going to "123" sub menu?

    I use Comma and question mark frequently at text messaging, email.....using my iPhone 4.
    Apparently, there's no short cut for this and I request Apple Support to create a shortcut for "," and "?' so there the Apple device can be more user friendly.
    Jackson

  • Calcscript Commit Block Interval

    What sets or adjusts the commit block intervals? On one calcscript I see messages saying "Commit Block Interval for the calculation is [3000]. On another when I run it, it says "Commit Blocks Interval was adjusted to be [7889] blocks. I am wondering how Essbase knows when to change the commit block interval? Is this something the calcscript would have done or is it something Essbase adjusts automatically? Or is it how the calcscript is written?

    Hi,
    Essbase automatically changes the value when running a calc script
    +"Essbase analyzes the value of Commit Blocks and Commit Rows during its analysis of feasibility for parallel calculation use. If Essbase finds the values set too low, it automatically increases them."+
    If you want to find out it in further detail, have a read of http://download.oracle.com/docs/cd/E12825_01/epm.111/esb_dbag/dstinteg.htm and section "Uncommited Access"
    Cheers
    John
    http://john-goodwin.blogspot.com/

  • A few short SD blocking questions...

    Hi all,
    i got a few questions which the most of you probably can answer rather easily:
    1. Is it possible in SD (5.0) to register that a customer has to be blocked on a future date?
    2. Is it possible in SD to block a customer only for ordering (read: not allowing for new orders, thus new open credits) and not for all other functions (so that the customer still is available for corrections like credit memos,etc.)?
    3. Is it possible to block vendors in batch (mass block)?
    Two more questions which are less important:
    4. My client doesnt do any archiving. The consequence of this is that there exist many customers in the system that are inactive. Is there a way to hide these customers (same for vendors) without deleting them or archiving them? Maybe i can give these customers the status "inactive"?
    5. Is it possible to give comment to all the changes that are in the customer master records so that an employee always can find back why certain customer information is modified?
    I really hope you guys can help me out in these question!
    Cheers,
    Wander

    Hi Wander,
    1. Changes to customer data are not time or date dependent. So it is not now possible to block customers on a future date.
    2. See transaction XD02 go to Extras -> Blocking data (or alternatively use transaction VD05) The blocking type determines what the customer is and is not allowed to do.
    3. Transaction MASS would spring to mind. field to change depends on the blocking type you want to set.
    4. I know you can mark a customer for deletion. I don't know what this does to the standard search helps.
    5. I guess a text field could help for this. But to my knowledge there is no standard solution for this.
    regards,
    Pim

  • Posting block Question in How to handle inventory management scenarios

    Hi all - I have a question in document
    <b>"How to...Handle Inventory Management Scenarios in BW"</b>
    can anyone please tell me what a posting block is in this document...
    also I did not understood what a Validity table is?
    thanks,
    Sabrina.

    Hi Sabrina!
    A 'posting block' period is something that now you are facing in inventory management, but you have to consider it in ALL logistic cockpit flows (what you can see in LBWE).
    This requirement can be easily explained in this way: as you know, to fill a setup table you have to activate the extract structure of the datasource for which you want to run the setup job (otherwise the system says that no active extract structure exists, do you remember ?); but, by doing so with the posting operations open (in other words, users can create new document or change the existing ones), you run the risk that these records are automatically included in the extraction queue (or in SM13 if you are using unserialized method) AND also collected in setup table ! And you will have the same records with the initial load (from setup table) and then the same from delta load (from the queue): a nice data duplication.
    To avoid this situation, a "posting block" (or a "free-posting period" or a "downtime") is requested. No one can post new document during your setup job...
    Hope now is clearer...
    Bye,
    Roberto

  • Commit Work Question

    I have a question for all of the guru's out there. I am updating multiple files in my Mobile Asset Process, but at some point I want to do a commit, but I only want to do the commit for one file not all of them. Is there a way to do that?
    Thanks
    Mike

    Michael,
    hope this will answer your question:
    In database update processes, two things happen:
    - the objects to be updated are locked
    - database changes are written to the database rollback segment
    Every work process or transaction runs in it's own logical unit of work (LUW).
    At end of this LUW, an implicit COMMIT WORK occurs committing the rollback segment to the database. From now on, database changes become visible for all processes and users. Also all functions called with the addition IN UPDATE TASK are released for processing in an update process. An explicit COMMIT WORK at the end of active program will do the same and additionally start all FORMs called with the addition ON UPDATE.
    In your case: A COMMIT WORK will commit <b>all</b> database requests as described above: There is no way to to 'selected' COMMIT.
    If you process too many objects, either the database rollback segment or the lock table will overflow.
    In terms of performance, it is better to wait for commit as long as possible because every commit means heavy processing for ABAP and database.
    It is good practice to do a COMMIT after a certain number of objects. Some prefer 100; my personal experience says 1000 is fine.
    ROLLBACK works the same way but it will clear the whole rollback segment making all changes undone. If you encounter an error after say 900 uncommited objects and a rollback is necessary, you have to start again right from the beginning.
    Hope this helps you to find the best way.
    Regards,
    Clemens

  • Java synchronized block question...

    If I have the following setup
    Hashtable h = ...; // there will only be one instance of h in the program's execution
    public void A(...) {
        syncronized(h) {
            // read or write from h
    public void B(...) {
        syncronized(h) {
            // read or write from h
    }I understand that only one of the synchronized blocks can execute at a time, however, does that also mean that each will be seeing the most recent copy of h? For instance, if I write something to h in A(), can I be guaranteed that when I read from h in B() that I will always read h directly from memory and not some cached value on that particular thread?
    Thanks!

    I was under the impression though that if thread 1 calls A() and adds something to h, then thread 2 calls B() and reads from h that thread 2 might have a different copy of h and not see the change that thread 1 made.
    I've been looking around and it seems that as long as you used synchronized as soon as the lock ends any changes to the object will be written to the main memory and also any time a lock starts the object is read from memory and not from that particular thread's cache (which may not contain the recent changes). Which I guess answers my original question, and Yes, each time one of those sync blocks starts they will have the most recent value of h, no matter what thread last modified it.

  • User-land async comms programming questions

    I've ported wx200d from linux code to solaris 7/8. It reads data on
    com1 from the RadioShack WX200 weather sensor hardware. The
    daemon compiles and runs but when it gets killed (any term signal) the
    serial port becomes unusable. Restart attempts fail with errors such as:
    "wx200d: bind: Address already in use" or "all ports busy".
    The C code sets up the async line with termios. Once the port is hosed
    trying 'fuser /dev/cua/a' sees nothing, so how do you detect what is
    hanging on the line or causing it to block further attempts to use it?
    How do you clear it? Is there a user-land command-line app that
    will flush/close/reset the serial port? Rebooting does not clear it.
    Do I have to do a 'pmadm -r -p zsmon -s ttya' to remove the zsmon
    port monitor? Should I be using /dev/term/a instead of /dev/cua/a if
    I am only opening the device O_RDONLY? I am frustrated in my
    attempts to find how-to's, programmer reference guides, tutorials,
    example code, whatever to help me understand impacts of SAF/SAC &c
    when programming Solaris 2.x async comms for other than modems
    or terminals (ie. Celeste's tutorial &c). I haven't had to deal with UNIX
    async comms since Interactive SVR3.2.2 when getty/inittab/uucp was
    simple. I have already rtfm'ed: pmadm,sacadm,ttymon,ttyadm,saf,sac...
    Pointers? Help? Explanations? TIA. Please Cc replies to:
    <[email protected]> or <[email protected]>
    regards, cal

    I guess my problem is that I have some basic college experience in digital circuit design, and would like the forms to be programmed and behave in the same fashion as digital logic.
    The "Clear Form" menu item, of course, would be equivalent to the  reset button.
    Perhaps it is possible to hook my code onto the "Clear Form" menu item?
    -Ramon

  • Another multi-record block question

    I have always struggled with this. I have a muti-record block(10 records displayed). Eventhough my block is a db block, I have not assigned it a table. Hence, it is a non-db block. If I set enabled on a txt box, it
    is enabling the entire column. I need to enable only the text box on which the cursor is/user navigated to.
    How to do it?
    Thanks,
    Chiru
    Edited by: Megastar_Chiru on Sep 21, 2010 2:03 PM

    ok..here is the explanation of the issue
    in my multi-record block, there is a drop down list column with 2 values "Monthly %" and "Half-Rent". So, for example, in record 1, the drop-down could be chosen as "Monthly %" while in rec2, it could be "Half-rent". Right next to this list column, I have a text box. When the user chooses "Monthly %", the text box need to be enabled to allow keying a % value (say 5%). While in rec2, since the user chooses "Half-rent", the text box needs to be disabled for this record. How to accomplish this?
    Thanks,
    Chiru

  • XMII v12 - JRA Action Blocks Question

    We are implementing v12 (AIX) and require ECC server load balancing for our production environment. Production environment we have multiple xMII servers and multiple ECC servers. (our development environment is standalone, single xMII server, single ECC server). I understand that the JRA action blocks should be used for ECC load balancing instead of the JCO action blocks.
    Could not find any information concerning the 3 choices available in the JRA Start Session action block. We do not know which deployed adapter to use.
    1. nonGlobalTx/eis
    2. shareable/eis
    3. unshareable/eis
    From a previous post, found the following link, but did not go into detail on the 3 adapter choices:
    http://help.sap.com/saphelp_erp2005/helpdata/en/6f/1bd5caa85b11d6b28500508b5d5211/content.htm

    Here's what it was supposed to say...
    David,
    Just for completeness on the earlier part of the thread here are the connection definitions:
    "Non-shareable" - means that inside a transaction 2 or more physical connections will be created if applications wants to acquire 2 or more connections from one resource adapter. Each connection which is acquired from application will use separate physical connection.
    "Shareable" -   means that inside a transaction 1 physical connection will be created even if the applications wants to acquire 2 or more connections from one resource adapter. All connections that are returned to the application will share one physical connection. This is the default configuration.
    Non-global TX - connections which are acquired from connection factory with this setting will not be enlisted into distributed(JTA) transactions. This is SAP specific setting.
    Global TX - Connections are enlisted into transaction. This is the default/standard configuration.
    As for SSO, I may have misspoken earlier about this still trying to find a good answer for you.  In the mean time here's the best that I can do:
    Application authorization - application is responsible to provide user/password during get connection request. Or  user/password is specified into resource adapter properties. This is the default configuration in SAP.
    Container authorization - means that server is responsible to provide user/psw based on specific configuration which is done on the server.  For example SSO is done with container authorization.
    The Container Authorization has a bunch of SAP Help ([http://help.sap.com|http://help.sap.com]) on the security configuration here:
    ([https://help.sap.com/javadocs/NW04S/current/jr/index.html|https://help.sap.com/javadocs/NW04S/current/jr/index.html])
    I will continue to follow-up on this but if you keep me updated on your progress it will help me along as well.
    Sam

  • Trivial Catch block question

    In a test harness I wanted to put some debugging output to System.out and keep one line per thread execution so I could read it more easily.
    I have a try catch block, and in both sections my debugging prints out with System.out.print, so they should appear on the same line. The strange thing I have noticed is that when ever the catch block is entered the code prints on a new line and I dont know why.
    Below is a snippet of code, its pretty simple but illustrates this:
    try {
       System.out.println("starting test - ");
       //at this line the test code runs
       System.out.println("success");  
    } catch(Exception e) {
       System.out.println("fail");
    }The output for a successful test is like:
    starting test success
    And for a fail:
    starting test -
    fail
    I am just curious to know the answer, its like a new line character is output when the print statement is within the try catch block.

    MarksmanKen wrote:
    You want to use System.out.print not System.out.println
    Notice the difference?
    One automatically starts a new line and one just carrys on the current line.Wrong: One automatically ends the current line, while the other one does not!
    A subtle difference, but important:
    This code will print "foo" and "bar" on two separate lines:
    System.out.println("foo");
    System.out.print("bar");This code will print "foobar" on one line:
    System.out.print("foo");
    System.out.println("bar");

  • " *COMMIT "  statement - Question

    *XDIM_MEMBERSET P_ACCT = CE0004010
    *XDIM_MEMBERSET TIME = 2006.DEC ,2006.Nov,2006.OCT
    [#CE0661000]  = 500
    *COMMIT
    In the above logic, how many records will get updated?
    I am guessing  the above logic will update 3 records(based on teh time dimension) , is that correct?

    Dear Sapki6,
    Sorry I did not clear your script logic. If you want to create a new record into BPC table, please you have to use *REC() statement and combine with *COMMIT() statement. *COMMIT() statement is used to write down a data into tables and maybe you want to view a new record, please you do not put *COMMIT() statement in your script and BPC will write down in memory. So, you can see log of data for a new record which BPC created it.
    Kind Regards,
    Wandi Sutandi

  • Pop Up Blocker Question

    When responding to the pop up blocker message the IPAD freezes and I cannot get it to "unfreeze".
    Turning the IPAD off and back on does not correct this issue - Please advise how this can be fixed.

    Is this in Safari or another app ? If Safari then try clearing its cache : Settings > Safari > Clear Cookies And Data (Clear Cache on iOS 4) and also Clear History
    If that doesn't work then try closing the app completely and then re-open it : from the home screen (i.e. not with Safari 'open' on-screen) double-click the home button to bring up the taskbar, then press and hold any of the apps on the taskbar for a couple of seconds or so until they start shaking, then press the '-' in the top left of the Safari app to close it, and touch any part of the screen above the taskbar so as to stop the shaking and close the taskbar.

Maybe you are looking for

  • Cannot open as app in Windows 8

    I've updated my Skype and am frustrated that I cannot use it as an app in my Windows 8.  When I tried to open the app it its only function was to tell me to download the new software.  I used to be able to drag the app to the side so I could view it

  • Exporting using Compressor failed, and other awekward issues with FCP

    Hi FCP gurus! I need your help! Everytime I try to export using compressor, my submission fails. I even try a sequence a exported as m2v on Monday, but today is not working. Compressor will some times say "Unable to view preview, FCP is busy". When I

  • Manual production order creation in MTO

    Dear collegues, I am stuck in a scenario: I have a finished item A. strategy 20 is used.I created sales order for A in plant 1000. MRP run in plant 1000 happens. Through special procurement key 40, system generates planned order in plant 2000. Planne

  • F.13 is a standard program for automatic clearing?

    f.13 is a standard program for automatic clearing? can i include plant code to the standard program.....and it is used for display clearing......need information?

  • Error opening MPEG-2 in FCP and QuickTime

    I have an MPEG-2 elementary stream file created with Digital Rapids Stream under Windows. I get an error, "file unrecognized" when attempting to open the file with QuickTime Pro 7 and Final Cut 5. The error code was -2048. I have the MPEG-2 player co