Replication :: Refresh_all_mviews Returning Failures But Deferror Is Empty
May 7, 2012
I have successfully cleaned up our test environment but am having problems with the development environment.
We have a master table, a single materialized view log on that table, and 85 materialized views based on them.
Following is the output from my 'refresh_all_mviews' and query on 'deferror':08:36:04 SQL> SET serverout ON
DECLARE
v_failures NUMBER(12) := 0;
begin
dbms_mview.refresh_all_mviews(v_failures,'F','', refresh_after_errors => true, atomic_refresh => false);
dbms_output.put_line('Failures='||v_failures);
END;
/
08:36:33 SQL> 08:36:33 2 08:36:33 3 08:36:33 4 08:36:33 5 08:36:33 6 08:36:33 7 08
:36:33 8
Failures=5
PL/SQL procedure successfully completed.
08:36:47 SQL> 08:36:47 SQL> select * from deferror;
no rows selected
08:36:56 SQL>
why no rows are being selected? Do I have to 'set' something 'on'?
Actually am trying to replicate two db servers from one in hong kong and another in china. when am trying to establish the replication, am getting error 'ORA-04052: error occurred when looking up remote object' like this...
but the same way i have tried in my local network, it is working fine.i have tried schema replication through enterprise manager grid control..
i need to set up a central server with all the master tables and two other local database which will hold the updatable materialized view of the master table...the databases must be synchronized with central server..and user will work on the materialized view database...
I want to set up advance replication for 3 master site (multimaster) I created 3 master site named orc1,orc2,orc3 and followed up oracle replication management of API book instruction I created 2 tables(tes1,test2) in hr schema in all 3 master site with the same data. then I created the following steps
1-CONNECT repadmin/repadmin@orc1
2-Create the master group named hr_test_repg
BEGIN DBMS_REPCAT.CREATE_MASTER_REPGROUP( gname => 'hr_test_repg'); END; /
4-add tables test1 and test2 to the group
BEGIN DBMS_REPCAT.CREATE_MASTER_REPOBJECT( gname => 'hr_test_repg', type => 'TABLE', oname => 'test1',
[code]....
I could create DBMS_REPCAT.GENERATE_REPLICATION_SUPPORT for test2 but not for test1 and it produces error
RROR at line 1: RA-23309: object hr.test1 of type TABLE exists RA-06512: at "SYS.DBMS_SYS_ERROR", line 105 RA-06512: at "SYS.DBMS_REPCAT_MAS", line 2552 RA-06512: at "SYS.DBMS_REPCAT", line 562 RA-06512: at line 2
I have one 10g database in other country. I want part of their db (selected tables or tablespaces) and import that data to my 10g DB and i want keep to date this data.
I know two ways
1. Data Pump Imp/Emp via FTP, but i can't send only data that have changed (incremental), i must pumping whole selected part of database (i want only new data from their DB, but consistanse with my DB)
2. RMAN etc. or other archivisation tool, i can do incremental achivisation, but can i send files to another instances (my db) and load only that data? Can i do that with SQL*Loader?
For using replication in our production, here i am testing golden gate as replication tool. I tested all scenario in Uni direction ( source to destination). Now have to test replication with DDL support in Bi-Direction. Not getting any Doc for doing replication in Bi-direction( Two Way). If any one has done the same, then please share limitation of replication in Bi-direction through Golden Gate.
I Have configured the replication between two database's at table level. After few miniuts of successful configuration of replication between two db's at table level, I am getting the ORA-00001: unique constraint (%s_PK) violated error in dba_apply_error.
I checked constraint name,type and status on table replicated is same on both source and destination db.
I am using prebuilt MV to perform replication of about 300-400 master tables from one database to another database. I am wondering about the impacts on triggers in general replication.
IS there a general rule to enable/disable a trigger before a refresh.
I want to configured A Synchronous replication in oracle 8i. Noe using GUI i am bale to create replication but whatever i modify it remain in the transaction i have to manually run the job ,
I also tried to configured continuous asynchronous configuration by setting delay rate 500000. but it was also not working.
Is it possible to replicate table data on real time from sql server (2005 32 bit or sql server 2000 32 bit)to oracle 10g running on linux 64 bit? If yes then what are the steps.
It will be one way replication from sql server to oracle. Which option is best sql server dts or Oracle Stream replication to replicate table data.
Once a year in the application we have a specific query that gets used a lot. It's an UPDATE that updates a single record in a single table with a few different datatypes, but the issue is happening with one of the VARCHAR2 fields. It updates one VARCHAR2(2000) and three VARCHAR2(4000) fields at the same time.
This year, 9 of the 95 times it was used resulted in one of the VARCHAR2(4000) fields as null in the database. The users would not want this field to be null and 5 of the 9 have told us they entered something (the form they're filling out is a research proposal and leaving this field empty would be pointless because it's part of the funding request, so they're not doing it). The application isn't doing it because it's not consistent. I've checked the application and these fields can't be nulled any other way.
We just found the issue so I looked back over the past years back to 2005. Last year it didn't happen at all. In 2010 it happened a handful of times. Some years there were even more times. It's not always the same field but it's always a VARCHAR2 of at least 2000 characters.
I have a lot more information but it's all just details (let me know if you need to know more). I'm wondering if there is a bug in 10g with these types of fields. I don't believe it's malicious behavior on an individual's part but I suppose that's always possible.
how to research something like this. I tried to get access to Oracle Support and the Knowledge Base I heard they have but it doesn't look like I can do that
I have a query that will either return one record or zero records. When it returns zero records I want to replace my attributes with a sentinel, like 'N/A'. I tried the CASE statement but couldn't get anything to work
Sample (does not work): select (case when exists (select product from tbl_product where productid = '123') then product else 'N/A' end) product from tbl_product where productid= '123'; If one record exists it should produce: 'My Widget' (or whatever) If zero records exist it should produce: 'N/A'
What is the reason, and It is saying that temporary tablespace is empty ( temp tablespace have almost free space ) why it is saying like that? And users default temp tablespace is that tablespace only ( database have only on temp tablespace
I used a dynamic item_list. This list should contain only the result of the query but I found an empty element in the list. So, how to eliminate this element empty list?
There is one table with data in ORCL1 database.I have created the table using create table statement in ORCL2 database.Now i want to insert only the data into table.
I know one method ... drop the table i created using drop table statement and then create the table with data using export/import.
Is there any other way we can load data into empty table?
My db version: Oracle 11g I have an empty csv file.I created a external table for the empty csv file.When I run:select count(*) from externaltblname;It returns 1. It should return 0 right. In the definition, I specified "SKIP 1"But still it returns 1. When I use this external table to load into a target table. It loads a single row with null values.How to fix this.
what empty blocks are, and how to remove them.What I'd like to do is not have empty blocks in the first place on loading a table. I load a lot of "static" tables and would like to not have any wasted space at the end, with minimal shinanigans.
I've set pctfree 0 I"ve set initial to close to the end table size I've set next to 1M I've set pctincrease 0 blocksize is 8k
Yet I still need to at least do an alter table deallocate unused