Improved performance of extraction of ASO database

Similar Questions

• Improving the performance of OLAP in Oracle Database 11

  Hi all

  First of all sorry for a very generic question, lately I been faced with some problems of performance during the generation cubes

  for example we had a problem of low SGA_memory, now we have referred questions to Doc-ID 1464064.1

  to cut short-we have a list of things to look for in advance for OLAP can improve performance?

  similar to the one mentioned in a blog post (improving the performance of OLAP in Oracle Database 10 g Release 1) which is for 10 g

  we ORDM DW, which has the dimension of TIME based with about 10 k records sometimes, that it takes about 20 minutes to build recently its not not building at all throwing error mentioned Doc-ID 1464064.1 I have a wip SR to even SR 3-9620792761

  Similarly, we produced Sun with approximately 9million records sometimes he finished in 4 hours sometimes it ends quite

  I have twice fallen & recreated workspace that helps a little but does not appear inconsistent.

  Pointers would be really useful

  Thank you

  Prateek

  You can update following anonymous pl/sql block and change in your

  specific information for the name of the cube and measure names. You

  can add more steps as you like following the same convention of xml.

  DECLARE

  CLOB xmlCLOB;

  tmp varchar2 (4000);

  BEGIN

  DBMS_LOB.CREATETEMPORARY (xmlCLOB, true);

  DBMS_LOB. Open (xmlCLOB, DBMS_LOB. LOB_READWRITE);

  tmp: ='

  ';

  DBMS_LOB. WriteAppend (xmlCLOB, length (tmp), tmp);

  dbms_cube.import_xml (xmlClob);

  DBMS_LOB. Close (xmlClob);

  END;

• To improve performance, increase the size of the overall storage cache

  I have an ASO in 11.1.2.1 on unix Box
  
  I get this message in my logs
  
  "To improve performance, increase the size of the cache of global storage.
  
  What is the best practice to update the Cache of ASO for dérirez best time
  
  Please notify

  Right-click on the database in the EA and select Edit-> properties. It is on the first tab, remember to restart the application so that they enter into force

• improve performance on WHERE "ABCDEFGHIJK" AS FIELD | '%'

  Hi all
  
  I have several tables with a constant number of cases (between 1000 and 1 million) and an indexed field (usually a primary key) I need mark from a PL/SQL function. The database is 11.2.0.2.0 on RedHat.
  To make it simple guess this:

  CREATE TABLE A (A VARCHAR2(20) CONSTRAINT A_PKEY PRIMARY KEY);
  INSERT INTO A VALUES ('ABCD');
  INSERT INTO A VALUES ('ABCDEF');
  INSERT INTO A VALUES ('ABCDEFG');
  INSERT INTO A VALUES ('ABCDEFGH');
  INSERT INTO A VALUES ('AACDE');
  INSERT INTO A VALUES ('AACDEFG');
  INSERT INTO A VALUES ('AACCDDD');
  -- between 1000 and 1 million inserts
  COMMIT;
  ANALYZE TABLE A COMPUTE STATISTICS;

  I need the file with the longest field value corresponding to the front of a particular value, for example "ABCDEFGHIJKLMN". The correct result is 'ABCDEFGH '.
  A simple solution is this:

  SELECT * FROM (
    SELECT * FROM A WHERE 'ABCDEFGHIJKLMN' LIKE A || '%' ORDER BY LENGTH(A) DESC
  ) WHERE ROWNUM = 1;

  The problem is that the DB is not efficiently using the index (EXPLAIN the PLAN said Index Full Table Scan or Full Table Scan).
  My real solution is to query from the longest possible values:

  DECLARE
     RESULT VARCHAR(20);
  BEGIN   
  FOR LEN_PREFIX IN REVERSE 4..LENGTH('ABCDEFGHIJKLMN') LOOP
      BEGIN
          SELECT A INTO RESULT FROM A WHERE A=SUBSTR('ABCDEFGHIJKLMN',1,LEN_PREFIX);
          EXIT;
      EXCEPTION        
          WHEN NO_DATA_FOUND THEN
              RESULT := NULL;
      END;
  END LOOP;
  -- Do something with RESULT (it can be NULL too)
  DBMS_OUTPUT.PUT_LINE('RESULT: ' || RESULT);
  END;

  Is to use the primary key well and the query is compiled (not dynamic, it is located in a Package + PL/SQL function).
  What I don't like is that I'm running several queries (end with one exception) before obtaining the final result of "ABCDEFGH".
  Since I was 50 million to different values of 'ABCDEFGHIJKLMN', to the query it is increase the number of queries per about 5 times (or more)!
  
  Perhaps another simple solution is to place all the values in a list and then apply the ORDER BY to get the longest value:

  SELECT MAX(A) KEEP (DENSE_RANK LAST ORDER BY LENGTH(A)) FROM A
  WHERE A IN ('ABCDEFGHIJKLMN', 'ABCDEFGHIJKLM', 'ABCDEFGHIJKL', ... , 'ABCD');

  Explain the Plan said SORTING (ITERATOR INLIST (INDEX UNIQUE SCAN))).
  Because I want a compiled query I need to replace this list in a nested table or other objects (not sure which is better).
  But this solution has the disadvantage that any Oracle internally makes a query for each possible value such as the explain Plan tells me. While with the LOOP FOR that above, I'm from the longest to the first found value.
  
  Here's the code using a nested table:

  CREATE TYPE A_TABLE_TYPE IS TABLE OF VARCHAR2(20);
  /
  DECLARE 
       ALIST A_TABLE_TYPE := A_TABLE_TYPE();
       RESULT VARCHAR(20);
  BEGIN
  FOR LEN_PREFIX IN 4..LENGTH('ABCDEFGHIJKLMN') LOOP
      ALIST.EXTEND;
      ALIST(LEN_PREFIX-3) := SUBSTR('ABCDEFGHIJKLMN',1,LEN_PREFIX);
  END LOOP;
  SELECT MAX(A) KEEP (DENSE_RANK LAST ORDER BY LENGTH(A)) INTO RESULT FROM A
  WHERE A IN (SELECT * FROM TABLE(ALIST));
  -- Do something with RESULT 
  DBMS_OUTPUT.PUT_LINE('RESULT: ' || RESULT);
  END;

  I don't know if a nested query is preferred instead of the DENSE_RANK (I need other fields in the record).

  SELECT * FROM (
    SELECT * FROM A WHERE A IN (SELECT * FROM TABLE(ALIST))) ORDER BY LENGTH(A) DESC
  ) WHERE ROWNUM=1;

  Can I add fields, indexes, or whatever is needed to improve performance. Not sure if a partition is useful since they are small tables in any case (although with the number of records).
  
  Any help is welcome,
  
  byeee
  Andrea
  
  Published by: 888953 on October 3, 2011 05:27

  with tst
  as (select substr('ABCDEFGHIJKLMN',1,4+rownum) val, rownum rn
        from dual
       connect by level<= (length('ABCDEFGHIJKLMN')-4)
      )
    select *
      from (select (select a
                     from a
                    where a=e.val ) res
              from (select val
                      from tst
                     order by rn desc
                   ) e
           )
     where res is not null
        and rownum<=1
  

  Edited by: user6806750 the 03.10.2011 01:29

• Data extract - source ASO to BSO target

  Hello
  
  Client wants to retrieve wound of members of a dimension sku # a Cube ASO to the BSO Cube. Excerpts from OSI to ASO data are normally quite simple with DATAEXPORT. However, since ASO cannot run traditional calc scripts I'm wondering if there are methods to extract data from ASO to OSB? ASO does not store the values t rollled - it? So if it not is not wouldn't that eliminate partitioning option? Also, does ASO sopport report Scripts? Curious as not a guru of the ASO.
  
  Thank you

  You can have an XREF on the OSB by extracting data from the database of the ASO, though I wouldn't have thought that would help and xrefs can be slow.
  Try other options:- Re: extraction of ASO cube data

  See you soon

  John
  http://John-Goodwin.blogspot.com/

• What is the best software to improve performance and free hard disk cleanup. I used MacKeeper, but I'm tired of paying annual fees for this service.

  What is the best software to improve performance and free hard drive cleanup? I used MacKeeper, but I'm tired of paying annual fees for this service.

  You shouldn't have these software.  See discussion of Klaus1

  Do not install MacKeeper

  See also this one by Klaus1

  Viruses, Trojans, Malware - and other aspects of Internet Security

• Cannot open a file error where zip is a "compressed (zipped) folders error". Windows cannot perform the extraction.

  Original title: loading with a password required

  I'm trying to open a zip via windows 7.  Same error occurs on windows xp.  There is a "compressed (zipped) folders error" error  Windows cannot perform the extraction.  The destination file could not be created.

  I recently got 8 deleted window and window 7 set in... I've never had a problem before

  Hello

  1. Using default Windows 7 application or any other 3rd party application?
  2. What type of file you are trying to open?
  3. You try to extract the zipped file documents? If so where are you trying to save?

  I suggest you try these methods and check the status of the issue.

  Method 1

  Try using Windows to extract the zipped files

  http://Windows.Microsoft.com/en-us/Windows7/compress-and-uncompress-files-ZIP-files

  Method 2
  Perform the check and the boot if the problem persists.

  How to troubleshoot a problem by performing a clean boot in Windows Vista or in Windows 7
  http://support.Microsoft.com/kb/929135

  Note: follow step 7 of section of boot KB929135 to the computer in normal mode.

• Aggregation in the ASO database view

  Hello world

  I am trying to create a new view of aggregation for the combination of members (Level0 and high level members) on the ASO database. Not sure to create the aggregation process.

  I poured the current view_file using the "validate a global selection on dump Appname.dbname database to view_file 'MYVIEW';'. ...

  But, how can I change this file in my combinations of members... I went through some documents, but I have found no solution for my problem...

  Could someone help me pls with this one... (With steps would be very useful)...

  Thanks in advance!

  For the better or the worse, Oracle did not have what you want (create a view for a specific set of levels) easy.  The only guaranteed way to force creation of a view to a particular set of levels is to reverse engineer the algorithm which translates integers found in the file .csc views.  I did, but I do not recommend it particularly.  Alex Pavel deserves to have a translator to work entirely online, here: http://essbase-ish.blogspot.com/2014/02/custom-built-oraclehyperion-essbase-aso.html

  I also had success with the following query "cheats", using advice outline etc..  More on these techniques in this 2011 Kscope (mine...) presentation: http://www.odtug.com/p/do/sd/topic=33&sid=709 (free associate member required).

• How ORA_HASH can improve performance?

  I want to know how ORA_HASH can improve performance.

  Thanks in advance

  Not sure I understand what you want.

  ORA_HASH is just a function of Oracle (often used for the sample data by providing a hash value).

  It's nothing to do with the performance as such.

• Improved performance in vCenter

  Hello all - I hope that the community can help out me with this one.  It is related to the performance of vCenter.

  VirtualCenter arrived at the point where it's VERY slow almost all the time and I cannot understand why.  As examples, in the computers view models, and virtual VC locks up often for anywhere from 20 to 60 seconds, just by selecting one data center to another.  RVTools almost 20 minutes to come.  Also, if I try to install VeeamONE or VeeamBackup, they are slow slow is not usable and sometimes lower VC (so Veeam is not running - I just wanted to give an example).  Finally, if I restart the server (VM) VirtualCenter, it can sit to 'Apply the settings of the computer' for about 20 minutes.

  I support a vSphere environment that consists of the following

  vCenter: 5.0 Update 1 - running on a virtual of Win2k8R2 computer with 8vCPUs and 20 GB of memory

  SQL Server VC: 2008R2 SQL - running on a VM Win2k8R2 with 8vCPUs and 32 GB of memory

  MUV: VM dedicated

  VC Web: VM dedicated

  * The four run VMs on a dedicated Cluster 2 nodes 'management' which consist of two HP DL380p G8s DualProcs & a lot of memory. No other virtual machines running on the host & loans are all practically zero.

  He currently manages 316 ESXi hosts (all 5.0 Update 1) and about 6000 virtual machines.  When VC was installed, I chose the "Large" scale that we will add more than about 200 guests & about 1000 VMs more next year.  The use of the Cisco Nexus1000v ESXi hosts (excluding the 2 cluster nodes mgmt - those that are on the standard vswitches).

  The hosts and virtual machines are in data centers about 20 & 30 clusters, physically dispersed across the country (US).

  Any thoughts on how to make it better?

  Thanks in advance!

  From there, should I run the script to purge old data from the other KB, or should I expect to see improved performance immediately?

  Yes, but... To do the purge your vCenter Server Service must be stopped.  Always better this offline.  We usually do maintenance on our vCenter DB about each quarter.

• How views can improve performance

  Hello

  Can someone explain to me how the views can improve the performance

  When DMLoperations perform through VIEW rather than perform DML operations by table.

  Thank you and best regards,

  Ibrahim Sayyed.

  84b09b47-0add-4854-b184-296b3903b009 wrote:

  Hi Ibrahim.

  A view is not much more than a macro which is developed by the SQL engine when running. Generally speaking, views will not improve performance.

  You can take the help of INDEXED VIEWS to improve performance. Then, views have clustered index assigned and, when they do, they will store the temporary results that can speed up queries that result. These views can speed up and simplify the application design.

  What is the INDEXED VIEWS?

• ReceiverThreads JMS connector in 10.1.3.3 layout to improve performance?

  Gurus,
  
  I use nonBlockingInvoke = true property in bpel.xml to spawn a thread for each parallel flow activity.
  
  However, the definition of this property I found success in performance. I suspect that the parameter of ReceiverThreads to be the question JMS connector. If ReceiverThreads more could hear in the queue, you can improve performance. (Page 47 of the 10.1.3.1. Performance Document).
  
  But I'm not able to locate this property in orion-ejb - jar.Xml. This property holds a different file or this property needs to be added (please can you provide the exact code)?
  
  Thanks a ton!
  SP

  You can find the documentation of ReceiverThreads here:
  http://download.Oracle.com/docs/CD/B32110_01/Web.1013/b28958/JMS.htm
  (Table 4 - 11).
  There is above the table of examples that show how to add a property.
  (A property can be added to orion-ejb-jar. XML or ejb - jar. Good XML if you set the property via a file .xml orion-ejb - jar.xml is recommended because it is a specific OC4J parameter).

  Alternatively, you can set via the annotation. See here:
  http://download.Oracle.com/docs/CD/B32110_01/Web.1013/b28221/actcfgprop.htm

• How can I improve performance Adobe CS5?

  I got a trial of Adobe Creative Suite 5 on a PC at work and so far it's ok, but it's really trolling. The PC currently has an Intel Core 2 Quad Q9400 2.66 GHz, 4 GB of Ram DDR2, 160 GB HDD, and an integrated Intel graphics chip.

  Add a dedicated graphics card will improve performance of CS5? I read on the forums that Mac users use a 2nd hard drive as a "scratch disk" for the likes of photoshop which improves performance but do not know how to implement that.

  Any ideas would be great.

  The integrated graphics may well take all system RAM. In addition, you enter your operating system but I recommend Win 7 64 bit with a boat load of RAM.

  Fast hard drives with a drive of dedicated work will help, too.

  Bob

• Super Proxy? Improve performance more?

  http://help.Adobe.com/en_US/AfterEffects/9.0/WSF13D6BED-c53b-408a-B2D6-C8B4205D4FB7a.html

  Apart from these tips from Adobe on improving performance, does anyone know any other tips?... even if it is a purchase of third party. It seemed that just when mandated, the machine still flat enough to annoy me and theyre proxy/lowres images is of course a bit blurry. I guess Im hoping there are some great plugin proxy that allows you to work superfast. The dream is to create animations 3D HD on a laptop. FTP then all in a cluster, and they do most of the work and send me a nice shiny bluray.

  No, sorry. There is no magic recipe. Computational mathematics defeated this dream. AE is a full buffer rendering engine, which means that everything that this process must adapts perfectly to the RAM. So, if your machine is not at the height - well, you know the story. That will change only when AE Gets a converter in mosaic or a user-friendly method to same memory or knows how to deal with a lot more RAM in a 64-bit version. So far, these solutions will have to get by.

  Mylenium

• Improved performance for EBS

  Hi all
  
  We currently have a machine for EBS R12 dual server configuration. One (1) server for applications and 1 for the DB.
  In my view, is a standard and the most common installation.
  Assuming that we have another 1 server just to improve performance, which component of EBS devrais I moved
  in the 3rd Server?
  
  
  Thank you very much

  It will be managed by balancing the load itself. As far as I KNOW, you have no control over this (i.e. you can tell the user which layer application for access to).