PS_TXN table - what is it?

I created a JSP page connected to a database, but when I change the login and password (of course authenticated to the database) I get the following errors:
JBO-30003: The application pool (OPTEN_MP.OPTEN_MPModule.OPTEN_MPModuleLocal) failed to checkout an application module due to the following exception:
oracle.jbo.PCollException: JBO-28033: Could not retrieve row from table PS_TXN, collection id 12 941, persistent id -1
java.sql.SQLException: ORA-00942: table or view does not exist
I haven't any table named PS_TXN
Thanx
Zoltan

It will create the table in the current schema (using a separate connection) unless an alternative is specified as explained in the link above.
The framework will create the necessary tables dynamically as long as the user connecting to a schema has create any table, create any sequence etc (again see link above).

Similar Messages

Index for PS_TXN table

Hi
This is regarding the PS_TXN table which is created and maintained by ADF.
We developed our application using JDev build 1811 and Oracle 10g as database.
We recently migrated to JDev build 1929. We have the BC4J cleanup job in our database to clean the PS_TXN table.
Since ADF is running a full table scan with each query, we have got a suggestion from our database administrator to create an index for the column COLLID in PS_TXN table.
My question is:
1) Since this is a JDeveloper table, is JDeveloper responsible for automatically providing indexes?
2) If this is true, why was the index suggested by Robert (COLLID) not automatically added?
4) If we manually add this index, is there a possibility that JDeveloper will override our change?
Thanks in advance
Shoba

Hi
Thanks for the immediate reply.
We have not set the property. But as per my knowledge, if the value of this configuration parameter is not set by the developer (which is the default situation) then the framework will create the temporary tables using the credentials of the current application database connection. Thats what happens in our applicaiton. It created the table in our schema and the cleanup job is also working fine.
But why an index is not created for the table PS_TXN by JDev?
If JDev does not create any index, can we create one for performance.
If we create, will it overriden by JDev.
Thanks in Advance
Shoba

Passivation and PS_TXN table - (Oracle ADF 11.1.1.6)

Hi All,
Can anyone help me understand the relationship between passivation and PS_TXN table insert? I am testing my application in a controlled environment using following steps -
(1) I truncated PS_TXN table first.
(2) I ran my application and visited couple of pages.
(3) I queried PS_TXN table and it has 22 rows in it
(4) When I check in enterprise manager ADF performance page the number of Passivations per min still shows 0 rows for all of my AMs.
What else could cause inserts in to PS_TXN table?
Thanks
Chetan

See My Oracle Support note IMPDP Fails With Errors ORA-39126 And ORA-01403: no data found (Doc ID 729346.1).
Note that is a forum and not official Oracle Support: you should not ask for urgent answer.

Bc4j and PS_TXN Table

Hi,
Reading the article "Understanding Application Module Pooling Concepts and Configuration Parameters" from Dive into Bc4j Blog; http://radio.weblogs.com/0118231/stories/2003/12/17/understandingApplicationModulePoolingConceptsAndConfigurationParameters.html
The jbo.passivationstore, parameter sets where we want to store Bc4j's Xml snapshots ( files on disk or Database ).
Bc4j AM, uses PS_TXN table to store it's states between sessions calls.
Querying this table at an dummy database schema just to see how many rows, i realize that it has almost 7.000 rows .
Application Module Pool does not have any parameter to configure a garbage collection over Ps_Txn or the disk file ?
I mean, as it is a dummy schema the application module pool, never will restore any state stored there, how it's works ? When can we clean all that rows ?
thanks.

Passivation store parameter controls all passivation, including what is written to PS_TXN when you opt for the (Default of) database passivation.
BC4J should clean up the snapshots when the AM is released in stateless mode, or when the session expires.
If sessions are killed, we supply a script in ./BC4J/bin/bc4jsession.sql to clean up the info in that table at whatever interval you would like to.

PS_TXN Table index

Have two applications that are built using JDeveloper 10.1.2.2 (Build 1929) that uses the PS_TXN Table. Would it be beneficial to add an Index to this Table to potentially increase response time?
The reason I ask is, if I do a count on this Table it takes about 10 seconds to return a count value of 7000+ records. Yet counts done on other Tables in the DB take about 1-2 seconds to return a value.
Thanks
Fran

I have reviewed this white paper - "Overview of Temporary Tables Created By BC4J" located here:
http://www.oracle.com/technology/products/jdev/htdocs/bc4j/bc4j_temp_tables.html
There is nothing specific about indexes on the PS_TXN table, so it looks like we just need to add an index and see if performance improves!
Does anyone have any experience with doing this? What were the results?
Thanks

While defining a columnar table, what is the purpose of column store type

Hi folks
I have two questions related to columnar table definition.
1. What is the purpose of column store type.
While defining a columnar table, what is the purpose of column store type (STRING ,CS_FIXEDSTRING,CS_INT etc) , when I define a table using the UI I see that the column is showing STRING but when I goto EXPORT SQL it does not show. Is this mandatory or optional ?
2.VARCHAR Vs. CHAR - In the UI when I create the table I do not see the CHAR option , but I do see lot of discussion where people are using CHAR for defining the columnar table. Not sure why UI dropdown does not show it. I also read that we should avoid using VARCHAR as those columns are not compressed, is that true, I thought the column store gives compression for all the columns. Are there certain columns which cannot be compressed .
Please let me know where I can find more information about these two questions.
Poonam

Hi Poonam
the CS_-data types are the data types that are used internally in the column store. They can be supplied but it is not at all required or recommended to do so.
SAP HANA will automatically use the correct CS_-data type for every SQL data type in your table definitions.
To be very clear about this: don't use the CS_-data types directly. Just stick to the SQL data types.
Concerning VARCHAR vs CHAR: fixed character data types are not supported anymore and don't show up anymore in the documentation.
I have no idea why you believe that VARCHAR columns are not compressed but this is just a myth.
create column table charcompr (fchar char(20), vchar varchar(20));
insert into charcompr (
    select lpad ('x', to_int (rand()*20), 'y'), null from objects cross join objects);
-- same data into both columns
update charcompr set vchar = fchar;
-- perform the delta merge and force a compression optimization
merge delta of charcompr;
update charcompr with parameters ('OPTIMIZE_COMPRESSION' ='FORCE');
-- check the memory requirements
select COLUMN_NAME, MEMORY_SIZE_IN_TOTAL, UNCOMPRESSED_SIZE, COUNT, DISTINCT_COUNT, COMPRESSION_TYPE
from m_cs_columns where table_name ='CHARCOMPR'
COLUMN_NAME    MEMORY_SIZE_IN_TOTAL    UNCOMPRESSED_SIZE   COUNT   DISTINCT_COUNT COMPRESSION_TYPE
FCHAR       3661                    70285738            6692569 20              RLE
VCHAR       3661                    70285738            6692569 20              RLE
We see: compression and memory requirements are the same for both fixed and variable character sizes.
- Lars

Jbo.passivationstore set to file, but PS_TXN table is still being created

I've an app with application modules all having jbo.passivationstore set to file, a la the below:
jbo.passivationstore="file" However, when i test this, the app creates the PS_TXN table and PS_TXN_SEQ sequence in the database. From reading the doc, I'd expect that only if jbo.passivationstore were set to database.
Is there some other setting that needs to be configured in order to make the passiviation handled outside of the database?
am using JDeveloper 11g.
Thanks for your help.

Timo, just tried it, that didn't work - PS_TXN is still being created.
i found that just deleting the names from the properties in the JDev UI had no effect. so i then literally set the attributes to empty strings as below:
<Database jbo.locking.mode="optimistic" jbo.txn_seq_name="" jbo.txn_table_name=""/>is there some other step i should take?
thanks for the quick response.
-george

PS_TXN table

Hi
Our application was developed using JDeveloper Build 1929 and Oracle 10g database.
We have three schemas in our application and all these schemas share the same PS_TXN.
My questions:
1. Should we truncate the PS_TXN table before deploying the code in the application server?
2. Will three schemas sharing the same PS_TXN table create any issues?
Thanks in Advance
Shoba

Maybe this can help:
http://www.oracle.com/technology/products/jdev/htdocs/bc4j/bc4j_temp_tables.html

Query regarding passivation and PS_TXN tables

Hi All ,
I am working on a read only Dashboard UI where the DB user has only read privileges.
The jbo.server.internal_connection uses the same DB connection to create PS_TXN & PS_TXN_SEQ tables which fails for obvious reasons & I get the error -
"Couldnot create persistence table PS_TXN_seq".
I have disabled passivation at all the VO levels and also disabled jbo.isSupportsPassivation to be false at the AMLocal level , but still I am getting this error.
I have also increased the initial AM pool size in bc4j.xcfg & Connection Pool at the weblogic server level to avoid snapshots been written to this table.
Is there any way I can prevent any interaction with this table as the client too is not interested in any kind of passivation to happen for the time being.
Thanks

Thanks for your reply Chris.
Chris Muir wrote:
You might be taking the wrong approach to solving this. Rather than disabling the AM pooling (which btw is not supported by Oracle) to the database, instead you can get ADF to passivate to file or memory of the app server.I am confused as to what exactly jbo.ampool.issupportspassivation = false does then ? I read on one of the blogs that its a viable use case for Programatic VOs ?
Also regarding passivating to file system as per http://download.oracle.com/docs/cd/E12839_01/web.1111/b31974/bcstatemgmt.htm#ADFFD1307 , its not really a recommended approach.. hence was not going for that.
Can you please throw some more light ?
Thanks

Why BC4J PS_TXN table , sometime may occur data lock situation ?

My appliation be develop in Jdev 10.1.2 with BC4J/JSP solution, but in production environment sometime when the application user request is heavy will occur the PS_TXN data lock situation and caused the application be hang and no any other user can access the application, so what situation wrong or event happen may caused the PS_TXN lock issue ? who have meet the situation and have the good idea to avoid the issue ?

Hi Nies......
If you want to Import Libraries to your target client then you need BASIS Consultant..........
For any report writer report we need a library.
Go to GR23 & specify the library name 1VK & press 'Enter'.
If it doesn't display......
Please check with your FI consultant as they are the users of the report writer reports and they only create the same.
there may be chance that this library has been deleted.
Is your system newly upgraded?
Solution..........
Library 1VK is SAP standard and exists in table T801K in client 000.
1. Create transport of table with library and transport to target client.
2. Or in target client, use transaction OKD3 to import the library 1VK from 000.
3. You can also use transaction GR29 to import individual libraries.
Thanks
Dhruv Malhotra

SQL query with multiple tables - what is the most efficient way?

Hello I am learning PL/SQL. I have a simple procedure where I need to find number of employees and departments per location as per user input of location_id.
I have 3 Tables:
LOCATIONS
location_id (pk)
location_name
DEPARTMENTS
department_id (pk)
location_id (fk)
department_name
EMPLOYEES
employee_id (pk)
department_id (fk)
employee_name
1 Location can have 0-MANY Departments
1 Employee has 1 Department
Here is the query I came up with for PL/SQL procedure:
/*Ecount, Dcount are NUMBER variables */
SELECT SUM (EmployeeCount), COUNT(DepartmentNumber)
     INTO Ecount, Dcount
     FROM
     (SELECT COUNT(employee_id) EmployeeCount, department_id DepartmentNumber
          FROM employees
          GROUP BY department_id
          HAVING department_id IN
                    (SELECT department_id
                    FROM departments
                    WHERE location_id = userInput));
I do get the correct result, but I am just wondering if my query is on the right track and if there is a more "efficient" way of doing this.
Thanks in advance for helping a newbie out.

Hi,
Welcome to the forum!
Something like this will be more efficient:
SELECT    COUNT (employee_id)               AS ECount
,       COUNT (DISTINCT department_id)     AS DCount
FROM       employees
WHERE       department_id IN (     SELECT     department_id
                    FROM      departments
                    WHERE      location_id = :userInput
;You should also try a join instead of the IN subquery.
For efficiency, do only the things you need to do.
For example, you don't need a count of employees in each department, so don't compute one. That means you won't need the in-line view, so don't have one.
You don't need PL/SQL for this job, so don't use PL/SQL if you don't have to. (I realize this question was out of context, so you may have good reasons for doing this in PL/SQL.)
Do all filtering as early as possible. Don't waste effort computing things that won't be used .
A particular example of this is: Never use a HAVING clause when you can use a WHERE clause. What's the difference between a WHERE clause and a HAVING clause? The WHERE clause is applied before aggregate functions are computed, and the HAVING clause is applied after; there's no other difference. Therefore, if the HAVING clause isn't referencing an aggregate function, it could be done in a WHERE clause instead.

What is VIEW in Database table what is the advantage

Hi:
I would like to get your opinion on what is the advantage of Table VIEW?
I read a note saying it is all about simplifying Query.
But when you come to XMLTYPE where there is only ONE ROW and NO COLUMN,
How do you create XMLVIEW for a large XML files?
Please help
Ali_02

Overview of Views
Read here...
http://download.oracle.com/docs/cd/E11882_01/server.112/e10713/schemaob.htm#i20690
XMLView or XMLTYPE ?
Overview of XML in Oracle Database
Read here...
http://download.oracle.com/docs/cd/E11882_01/server.112/e10713/cncptdev.htm#CNCPT1535
Edited by: ordba on Mar 4, 2010 9:50 AM

SELECT statement INTO TABLE - what determines order of entries in in.table?

Hello everyone.
I habe a question: A select statement selects multiple entries from a data base table that are put into an internal table.
The question : What determines the order of the entries in the internal table?
And what might change the order all of a suddden in that internal table?
Could reorg. activities/archiving (even ongoing) cause any different result in the order as before ?
I do not mean different entries but I rather refer to the sorting of the internal table, if it is not explicitly stated in the select or in parts of the coding.
Any explaination is appreciated!
Thanks!
CN.

Hi,
Its not the order in which you specify the fields in the where clause that affects the sequence of the selection.
What I said is that by default the entries are sorted by the primary key of the table.
I did a bit of a test out of curiosity and after seeing so many different replies on the thread. I was surprised to see the results of the different scenarios.
REPORT Z_TEST_SELECTION                        .
tables vbap.
data it_vbap type table of vbap.
data it_vbak type table of vbak.
select-options s_vbeln for vbap-vbeln.
select-options s_posnr for vbap-posnr.
"======================================================================
"s_vbeln has following entries
" 5
" 6
" 1
" 2
"s_posnr has following entries.
" 10
" Sorts in ascending order by VBELN and POSNR by default
select * up to 20 rows from vbap into table it_vbap.
refresh it_vbap.
"Below two cases are absolutely unreliable and I would suggest that a
"sort should be mentioned after these statements. So when we use select
"options specify values with the in clause, the sort is not reliable.
"Strangely , neither the sequence is not determined by the order in
"which the data was entered in the select option nor in the sequence of
"the primary key. And this is still a mystery to me about the order in which
"the entries are selected.
" Any answeres anyone ?
"1.
select * from vbap into table it_vbap
where vbeln in ('0000000002','0000000006','0000000005','0000000001')
and posnr in s_posnr.
* Resulted in
*900 |0000000001|000010|
*900 |0000000005|000010|
*900 |0000000006|000010|
*900 |0000000002|000010|
"=========================================================
"2.
select * from vbap into table it_vbap
where posnr in s_posnr
and    vbeln in s_vbeln.
* Resulted in
*900 |0000000002|000010|
*900 |0000000001|000010|
*900 |0000000005|000010|
*900 |0000000006|000010|
refresh it_vbap.
"Here the orders were selected in Decending order of Sales order, however
"the posnr is not mentioned in the order by clause, hence posnr was
"still in ascending order.
select * from vbap into table it_vbap
where posnr in s_posnr and vbeln in s_vbeln
order by vbeln descending.
write : 'done!'.
refresh it_vbap.
"Here the orders were selected in Descending order of Sales order.
select * from vbak into table it_vbak where vbeln in s_vbeln
order by vbeln descending.
" The entries were selected in the ascending order of the key VBELN and
" POSNR, so it seems that for all entries does not affect the selection of records.
select * from vbap into table it_vbap
for all entries in it_vbak
where vbeln = it_vbak-vbeln
and   posnr in s_posnr.
Of course if you change the order of the primary key (in the transparent table, not in the where clause), the entries will be sorted as per the changed key.
But we seldom change the the sequence of the primary key. In most cases we normally add a new field and make it a key field in addition to the existing key fields.
I hope this helps you.
There might be many more cases and examples than given above, which you can try and reply on this post. Lets see what we come up with.
regards,
Advait

SRM table CFF_CONT question : huge table, what is that table for ?

Hello
We are looking to have more details about table CFF_CONT on a SRM system
The tolat size of DB is 344 GB
and only for table CFF_CONT we do have 284 GB, compared to total size this is really huge
the rest of the DB is only about 60 GB
What is that table for ?
Would it be normal to see that table on SRM growing much more than the others ?
Is that table something that can be clean up ?
Those questions are poping up becasue we are doing a system copy with sapisnt and R3load
and we see that this table only is requiring a so huge amount of thime to be exported
Regards

By the way ...
That CFF_CONT table belongs to cFolder components

Pasting cells into tables - what has changed?

Sometime this year something has changed with the ability to copy and paste multiple cells from a table in a Word.doc into a table in InDesign CC. I used to be able to select the same number of cells, copy and paste directly then proceed with formatting. However now If I highlight and copy 12 cells in a Word table then highlight and paste 12 cells in an INDD table, all the 12 cells copy into cell one only. Is anyone able to shed some light on any change I should know about or different way I should be copying into INDD tables? It's a ridiculously time consuming way to be working on multiple tables.

Thanks for your help Laubender
I have followed your instruction, however now I have a new issue. Since changing to 'text only' in preferences, it breaks up every single line in the copied cell and creates new and unwanted cells in the INDD table.
So to reiterate, I have a table in word with 6 cells highlighted and information copied. I open up a new table in INDD with 6 cells highlighted. I paste into the 6 cells with no formatting and end up with this:
In one of the CC upgrades, a default setting has changed somewhere along the way as I never had this issue when using CS6 last year. I need to pinpoint EXACTLY what the setting is to return it back to normal.
Any other ideas?

PS_TXN table - what is it?

Similar Messages

Maybe you are looking for