Replication/ Master Site

Hi,
I4m trying to do a Materialized view site configuration and when i am finishing returned the following error ORA-02082 a loopback database link must have a connection qualifier.
Ok, i know error mean that: This was made to create a database link with the same name as the current database.
what i need to know ?
which are the necessary step to build master sites union two database table ?

Hello,
thanks for your answer:
the database names are the same.
We made already a full database export und imported it into the target database.
This was not succesfull, because after the import in the target database
the table dba_registered_snapshot was not filled!
Now: what to do?
Kind regards.
Steffen

Similar Messages

Migrating the replication-master-site-database to another hardware

Hello folks, hello oracle
we have got the followiong configuration:
--> Database 8.1.7.0
--> Advanced Replication system
--> Single master, snapshot configuration
--> 50 Slave-sides; offline replication once a day
--> per slave side: 50 RO-Snaps, 4 UP-Snaps
an the following problem:
we need to migrate the master database from
an IBM-AIX-host-computer to an HP-UX-host-computer.
the question is:
Wich steps do we have to make?
Is there a documented way or method to do this?
Kind regards
Steffen

Hello,
thanks for your answer:
the database names are the same.
We made already a full database export und imported it into the target database.
This was not succesfull, because after the import in the target database
the table dba_registered_snapshot was not filled!
Now: what to do?
Kind regards.
Steffen

Multi Master Site replication with Snapshot features

Dear,
I am implementing a Multi Master site replication. I want to replicate Table A in MASTER SITE A to Table A in MASTER SITE B.
However, I don't want all the data from MASTER SITE A to be propagated to MASTER SITE B.I may want data that belongs to a certain criteria only to be replicated over.
I know I can achieve this in a snapshot site replication. But can this be done in a Multi Master Site replication too ?

Hai,
As far under my observation, the tables that is marked for a replication from the MASTER SITE to Child Site is an exact replica of data. Your case could be achieved only thru SNAPSHOT VIEWS/GROUPS
- [email protected]

Error in Master Site Setup for Oracle Replication

Hi Guys,
While i was trying to setup the Master Site for Oralce 8.1.7
replication, a error is poping up like this :
ORA-06550:LINE 2, CLOUMN 4:
PLS-00201: IDENTIFIER 'DBMS_REPCAT_ADMIN.GRANT_ADMIN_ANY_SCHEMA'
MUST BE DECLARED
ORA-06550: LINE2, CLOUMN 4:
PL/SQL: STATEMENT IGNORED
looking forward to your kind help. If you have any idea about
solving this error then do contact me at my personal email :
[email protected]
Thanks in advince.
Souren

Niels,
1st check the storage prerequisites: http://itnewscast.com/chapter-6-oracle-vm-22-san-iscsi-and-nfs-backend-storage-configurations
2nd clean each servers cluster configs as show in the prerequisites section then create the pool.
Respectfully,
Roddy

Is it possible to create a super-set materialized view from master site?

In the replication documents, materialized view is a sub-set of master site. But my scenario is when I delete some records in the master site. I don't want these records(which's been replicated to the materialized view table before) be purged in the materialized view. Is it possible for oracle replication ?

If your goal is to replicate inserts and updates but to ignore deletes, you would probably be better off setting up Streams and simply not subscribing to deletes.
Justin

Missing Master Site in Multi Master Environment

Hello,
we are using MM-Replication on three Master sites with two replication groups.
Replication support was installed and configured by the application vendor.
I do not have experience with Multi Master Replication.
The data of the RG (Readonlymastergroup) was never replicated to one Master
(DEFS01). When I queried dba_repgroup on the failing site, I found the
RG status to be quiesced. There were no pending administrative requests
on this site or the Master Definition site, but I found that there is no definition
of the failing site's membership to the RG on the MD site.
Master Definition Site
select gname,dblink,Master,masterdef from dba_repsites;
GNAME DBLINK M M-----
MASTERGROUP      DEMD01 Y Y
MASTERGROUP      DEGS01 Y N
READONLYMASTERGROUP      DEMD01 Y Y
READONLYMASTERGROUP      DEGS01 Y N
MASTERGROUP      DEFS01 Y N
Failing Master Site
select gname,dblink,Master,masterdef from dba_repsites@defs01;
GNAME DBLINK M M-----
MASTERGROUP      DEFS01 Y N
MASTERGROUP      DEMD01 Y Y
READONLYMASTERGROUP      DEFS01 Y N
READONLYMASTERGROUP      DEMD01 Y Y
MASTERGROUP      DEGS01 Y N
READONLYMASTERGROUP      DEGS01 Y N
Can anybody explain how this could happen? AFAIK adding a master to a resource
group is a distributed transaction that should be rolled back on all sites, if it fails
on one.
To correct this situation; I am thinking of removing the RG from DEFS01
with DBMS_REPCAT.DROP_MASTER_REPGROUP (on DEFS01) and then rejoin DEFS01
with DBMS_REPCAT.DROP_MASTER_REPGROUP on the Master
Definition Site.
Will this work? Anything else I have to think off?
Regards,
uwe

Hi Janos,
I have tried the Multi language scenario. You need to have following setup in your system, they are:-
>Install language pack
>You need to upload content using below CSV fles:-
multistring.csv
Material.csv
Plant CSV
Product_Plant_Relationship_template.csv
If you can share your email id I will email you sample CSV file which you can use. Or else pl follow below steps to download CSV file from system:-
Click Setup
Click System Administration tab
Click to Import data
Click to New button
select radio button Upload to Server and click next
Select any dummy excel file from local server while browsing for Upload Import file option and click next.
Select Preview import check box adn search MultiString from the drop down object type and click next
In the page you will see the template.csv link, click and save the file. This the file you are looking for.
Create the content and upload the file...hope this will help you!!!
There is another way of importing master data is multiple langaure. In this scenario master data source would be ECC. You need to run SAP provided standard report to export the details. This XML file you can import into sourcing system. Pl find below blog link for more detail:-
http://scn.sap.com/community/sourcing/blog/2011/09/29/extracting-erp-master-data-for-sap-sourcing
Regards,
Deepak

Snapshot/master site implementation in one single system

how to setup Snapshot/Master site implementation in one single system. Can I get Sample code. Oracle 8 documentation has some problems.

I know what u said(that is understood)?
To get some practice on replication scenerio can i do it in single system by create two databases. one database as master site and the other as snapshot site. Is it possible on a single system.i'm able to create snapshot and master site in two databases but when i am trying to
DBMS_REPCAT.CREATE_SNAPSHOT_REPOBJECT () i'm getting problem using source given in oracle 8i document.
Thank u

Can I have more then one MVIEW site connected to single master site?

Hi,
I need one help from your end.
Currently I am working on Basic Replication.For that I am using Read only Materialized views.first i tried with one master site & one Materialized view site its working fine,Can I have more then one MVIEW site connected to single master site referring to same database ? if so how to achive this one..

Actually one of my senior told me that you need to change number of OPEN_LINKS value like that so for that reason I asked this question.
Still that problem not solved..
I will explain what are all the things I did ok.
IN MASTER SITE:
I am creating read only MVIEW's.
first i created one user repadmin and have given
1>execute any procedure
2>dbms_repcat_admin.grant_admin_any_repgroup
3>dbms_repcat_admin.grant_admin_any_schema
4>select any dictionary
and all other grants to repadmin user
then I changed global_name = 'zxincmp' on Master site.
Then I created mview log for each table (for 85 tables) & then I created mview's for that table on Master site itself. later I took a dump of these 85 mview's . then I transfered this dump to one of my mview site.
IN MVIEW site:
I created one user zxdbp_55 and all the above rights are given to this user & i created one public dblink
"zxincmp" using "TO_CMP"(master site).Then I imported that dump which i took from master site.
then i created one private link zxincmp in one of my mview site.then i used DBMS_REFRESH.MAKE('')
to refresh all my tables & to set the interval..
after that i created index on the mviews..
this will works out fine for single master & single mvirew site.but I tried with the same procedure to prepare one more mview site,everything is fine but main problem is second mview site is not getting refreshed automatically,i need to do refresh manually from enterprise manager.. then it will get refreshed.
how to overcome this problem? I am using same db name ,dblink,refresh group for both the mview site will it affects anywhere?
Thanks & Regards,
Hanumanth.

Can not refresh snapshot changes after importing data of master site

Hello !
I have two database computer,one as master site,one as snapshot site.Because the error of the hard disk of master computer,I use the exporting data file to recover my database.after importing ,I found I can't refresh the refreshgroup on snapshot,who can tell me why?
thinks in advance!
(exp system/manager full=y inctype=complete file='/home/save/backdata/xhsdcomp.dat')
(imp system/manager inctype=system full=Y file='/home/save/backdata/xhsdcomp.dat'
imp system/manager inctype=restore full=Y file='/home/save/backdata/xhsdcomp.dat')
null

You haven't listed the errors that you're receiving when attempting to refresh your refresh group, but if your snapshots are attempting to fast refresh, I suspect it's because the creation timestamp of the snapshot log on the master site is newer than the creation timestamp of the snapshot. In this case you will need to do a complete refresh of the snapshot (or drop and recreate the snapshot) before you will be able to fash refresh it again.
If this is not the case, please post the errors you are receiving when you attempt to refresh the refresh group.
HTH,
-- Anita
Oracle Support Services
null

The master site control file does not contain a component item for SMS_STATUS_MANAGER.

New to the System Center world. We are using a fresh install of Server 2012 R2 and System Center R2 with everything working until Server 2012 R2 Update 1 was installed. Now we are getting the following:
Site type: Primary Site or CAS SMS_STATUS_MANAGER
6/4/2014 1:35:21 PM 10996 (0x2AF4)
Resolved the "Status Manager" inbox to "C:\Program Files\Microsoft Configuration Manager\inboxes\statmgr.box\statmsgs".
SMS_STATUS_MANAGER 6/4/2014 1:35:21 PM
10996 (0x2AF4)
Opened a change notification handle for directory "C:\Program Files\Microsoft Configuration Manager\inboxes\statmgr.box\statmsgs".
SMS_STATUS_MANAGER 6/4/2014 1:35:21 PM
10996 (0x2AF4)
Resolved the "Site Control Manager (Master Site Control File)" inbox to "C:\Program Files\Microsoft Configuration Manager\inboxes\sitectrl.box".
SMS_STATUS_MANAGER 6/4/2014 1:35:21 PM
10996 (0x2AF4)
Opened a change notification handle for directory "C:\Program Files\Microsoft Configuration Manager\inboxes\sitectrl.box".
SMS_STATUS_MANAGER 6/4/2014 1:35:21 PM
10996 (0x2AF4)
Parsed the master site control file, serial number 1551209029.
SMS_STATUS_MANAGER 6/4/2014 1:35:21 PM
10996 (0x2AF4)
This site is the central site. SMS_STATUS_MANAGER
6/4/2014 1:35:21 PM 10996 (0x2AF4)
ERROR: The master site control file does not contain a component item for SMS_STATUS_MANAGER.
SMS_STATUS_MANAGER 6/4/2014 1:35:21 PM
10996 (0x2AF4)
Sleeping for 60 seconds... SMS_STATUS_MANAGER
6/4/2014 1:35:21 PM 10996 (0x2AF4)
Any thoughts on how to fix this? Seen a few references to a "site reset" for SCCM 2007 but wanted to make sure there wasn't anything else to try before heading down that road with 2012R2.

A site reset can also be done in CM12 so it's worth a try.
Torsten Meringer | http://www.mssccmfaq.de

Question on materialized view unregistered on master site

Hi
not sure how it is known, on master site, about the existence of an materialized view on another site referring some table of a master site. it seems registering is not necessary.
on master site orcl1 : I have a table A, and a materialized view log defined on table A.
on another site orcl2 : have a db link to orcl1 and a fast refreshable materialized view SCOTT.A_MVW (create materialized view A_MVW refresh fast as select * from A@orcl1)
on orcl1 I perform : exec DBMS_MVIEW.UNREGISTER_MVIEW (mviewowner => 'SCOTT' ,mviewname => 'A_MVW' ,mviewsite => 'ORCL2');
as a result the materialized view seems to be successfully unregistered (as it doesn't appear anymore in : select * from DBA_REGISTERED_MVIEWS; )
But, to my surprise, I can still perform fast refreshes on A_MVW, which also remove lines from the materialized view log . so how is site orcl1 still aware about the existence of the materialized view ?
it seems that, registered or not at master site, a materialized view behaves the same...or it doesn't ? where is it stored , on master site, information about materialized views (especially those fast refreshable) that reffer local tables ?
Thank you

The view allows the Administrator at the Master (Source) site to see information about MVs in remote databases that are querying his/her database.
A database in Singapore may be administered by DBA "Hemant" . This database may have built (with appropriate permissions !!!) an MV that queries table(s) in a database in London managed by "Alex". Alex can query DBA_REGISTERED_MVIEWS in his (London) database to see that a remote MV is referencing tables locally.
Why would "Alex" UNREGISTER the MV ? There's no real need to. Unless the information needs to be "hidden" from, say, an Outsource DBA who will be taking "Alex"'s place ! (just kidding !)
Hemant K Chitale

MASTER SITE에서 SNAPSHOT을 찾는 방법(ORACLE8)

제품 : ORACLE SERVER
작성날짜 : 2002-05-13
MASTER SITE에서 SNAPSHOT을 찾는방법
==================================
Explanation
오라클 8.0 이전버젼에서는 마스터 사이트는 스냅샷이
fast refresh를 수행하기에 충분한 정보를 가지고 있어야
한다. 마스터 사이트의 관리 ( 예 : 마스터 테이블의
drop 이나 recreate ) 를 위해서는 전체 분산 데이터베이스
환경에서 어떤 스냅샷들이 마스터 테이블을 참조하는지
알아야만 하는 어려움이 있었다.
오라클 8 이전 버전에서는 마스터 사이트는 원격지에 위치한
스냅샷에 대한 정보를 가지고 있지 않았다. 하지만
오라클 8에서는 마스터 사이트서에 스냅샷에 관한 정보를
등록해서 사용한다.
이를 위해 DBA_REGISTERED_SNAPSHOTS 라는 뷰가 새로 생겼다.
이 뷰는 일반적으로 DBA_SNAOSHOT_LOGS 라는 뷰와 join을 해서
사용한다.
SQL> DESCRIBE DBA_REGISTERED_SNAPSHOTS
Name Null? Type
OWNER NOT NULL VARCHAR2(30)
NAME NOT NULL VARCHAR2(30)
SNAPSHOT_SITE NOT NULL VARCHAR2(128)
CAN_USE_LOG VARCHAR2(3)
UPDATABLE VARCHAR2(3)
REFRESH_METHOD VARCHAR2(11)
SNAPSHOT_ID NUMBER(38)
VERSION VARCHAR2(17)
QUERY_TXT LONG
SQL> DESCRIBE DBA_SNAPSHOT_LOGS
Name Null? Type
LOG_OWNER NOT NULL VARCHAR2(30)
MASTER NOT NULL VARCHAR2(30)
LOG_TABLE NOT NULL VARCHAR2(30)
LOG_TRIGGER VARCHAR2(30)
ROWIDS VARCHAR2(3)
PRIMARY_KEY VARCHAR2(3)
FILTER_COLUMNS VARCHAR2(3)
CURRENT_SNAPSHOTS DATE
SNAPSHOT_ID NUMBER(38)
Example: 다음은 scott user의 orders table에 refresh해야할 2개의 스냅샷이 있다.
SQL> SELECT LOG_OWNER, MASTER, LOG_TABLE FROM DBA_SNAPSHOT_LOGS
WHERE MASTER = 'ORDERS';
LOG_OWNER MASTER LOG_TABLE
SCOTT ORDERS MLOG$_ORDERS
SCOTT ORDERS MLOG$_ORDERS
위의 결과 알게된 2개의 snapshot에대한 정보를 다음의 sql로 알 수 있다.
SQL> select owner, name, snapshot_site
from dba_registered_snapshots, dba_snapshot_logs
where dba_registered_snapshots.snapshot_id = dba_snapshot_logs.snapshot_id
and dba_snapshot_logs.master='ORDERS';
OWNER NAME SNAPSHOT_SITE
SCOTT ORDERS V804.WORLD
SCOTT SNAP_ORDERS NEGRIL.WORLD
오라클 7 마스터 사이트에서는 스냅샷을 등록할 수 없다.
오라클 7의 스냅샷은 오라클 8쪽 마스터에 자동적으로 등록되지
못한다.
만약 DBA_SNAOSHOT_LOGS에 MASTER='ORDERS' 에 해당하는
3개의 entry가 있을 경우나, 앞의 query 결과는 2개의 entry라면
오라클 7에서 스냅샷을 사용하던 사이트가 하나 사용되고 있을
것이며, 오라클 8쪽에 등록되지 않은 경우일 것이다.
Reference Ducumment
<Note 68707.1>

Merge replication - multi site, multi distributor, publisher, subscribers ...

Hi,
We have 2 sites at the moment. Each site has 2 servers. To reduce network load I was aiming at doing the following with a merge replication :
* configure one server at each site to be distributor and publisher
* configure the other server at each site to be a subscriber
* configure the 2 publishers to also be subscribers to each other to sync the sites.
Or even better:
* have all the servers be distributors, publishers and subscribers, so that if one dies, who cares as they all have the same configuration.
Also, I was thinking of:
* having the distributor and publishers use a DFS share replicated on both sites (or maybe the same share on each sql server ? not sure what the best practice is on this one)
However, I can't seem to find a good post / thread where it describes this kind of scenario in a step by step way nor what happens to the identity range management.
If you could share any light on this or point me to the right English terms to describe the above scenario, thanks in advance.
Olivier
PS: We only have STD edition, so peer to peer and AG are not available to US. (which is why we are using merge)
PS2: One way I could see this work is if I manually set the identity columns in the tables so that they won't overlap and make the range management manual. Then just setup publications / distributors and subscribers on each node. (But if I can dodge the
manual management, it be great)

Hi Brandon,
Thx for the reply
we are using web synchronization.
According to the documentation ('The business logic handler you specify is executed for every row that is synchronized.') I would think this will do exactly the same as our trigger on the article.
When we would use a business logic handler to update the record which is just uploaded to the Publisher I think we will end up with the same checksum error.
We worked around the problem by not changing the status of orders once they are uploaded to the Publisher.
For that part of the process we now have only 2 statusses anymore :
- Not to be sent to central server
- Upload(ed) to central server.
So an order is created at the subscriber and has the status 'not to be sent to central server'
In a synchro-window the user can mark orders which must be uploaded to the central server.
Only when the users clicks on the synchro button the orders which are marked (on screen) to be synchronised will get the status 'Upload(ed) to central server' and the pull subscriptions is started.
On the central server we process only the articles with the status 'Upload(ed) to central server'.
Before we had the statusses :
- Not to be sent to central server
- Upload to central server.
- Uploaded to central server.
And we had our own trigger on the Publisher which updated the status of records from 'Upload to central server ' to 'uploaded to central server' which was not reliable because we often had checksum errors on that article.

Will removal of Enterprice CA break AD replication between sites?

Hello.
I have a AD environment that have a CA on a failing Domain Controller. This server is scheduled for decommission, and are running a CA for the Domain. This server has issued certificates to the domain controllers from the Domain Controllers template.
I have no other use for this CA other than for Domain Replication. Based on this I want to remove the CA role completely from the domain.
If I follow this procedure:
http://support.microsoft.com/kb/889250
Will the domain replication break or still be in operation?
Regards
Tommy Rasmussen

If you require certificates for AD e-mail replication I would not recommend this procedure as all certificates would be revoked following this checklist (step 1). If you decommission a PKI but want its certificates to remain valid you would not revoke the
certificates but only create the long-lived CRL (steps 2 and 3). And you would need to make sure that new DCs get new certificates.
Are you really using AD *e-mail* replication? I am just asking because I often DC certificates are deployed automatically but not actually really used. Certificates are not required for default AD replication.
Assuming that certificates are required now (and will be required in the future) I would recommend instead:
Option 1 - new PKI before retiring the old one.
Configure a replacement PKI, make sure that the DC replication template is published at this CA, make all DCs get renewed certificates from the new CA, then decommission the old one.
On principle, the existing CA could also be migrated to a new server with another name, but handling the CDP and AIA URLs gets a bit messy if the default names have been used when setting up this CA (as these point to the existing server or an LDAP object
that has the same name as the existing server). So if the CA is only used for issuing DC certificates, I would rather create a new one.
Option 2 - new PKI after retiring the old one.
This would be an option if you don't plan to add new DCs soon:
Make sure all DCs have valid certificates issued by the existing PKI. Issue the long-lived CRL but don't revoke the certificates. Uninstall the CA service - the objects required to validate certificates will remain in AD. Details may depend on customizations
of the CDP and AIA URLs. If you used the default settings you might have also an HTTP URL pointing to the CA server itself - so the DNS record would need to point to a replacement server holding the CRL and CRT files.
This would work as long as you don't need new certificates - thus as long as the existing ones are still valid and you don't join new DCs to the domain. So you should perhaps setup another CA in the next months.
Elke

DEFERRED TRANSACTION QUEUE의 내용을 지우는 여러가지 방법 (replication)

제품 : ORACLE SERVER
작성날짜 : 2004-08-13
PURPOSE
advanced replication 환경에서, 특정 master site나 updatable snapshot
site은 다른 remote의 master site로 데이타를 propagation시키기 위해서
해당 local site에 deferred transaction queue를 유지한다.
remote site로 잘 전달된 데이타는 주기적으로 dbms_defer_sys.purge job에
의해 deferred transaction queue인 DEFCALL에서 지워지게 되는데, 경우에
따라 문제가 발생하면서 DEFCALL의 내용이 remote site로 전달이 안되면서
계속해서 지워지지 않고 남게 될 수 있다.
이러한 경우 다음 트랜잭션의 진행이나 전달에도 방해가 될 수 있어
강제로 지우고자 하는 상황이 발생하는데 그러한 경우의 조치 방법에 대해서
자세히 설명한다.
SCOPE
Advanced Replication Feature는 8~10g Standard Edition에서는 지원하지
않는다.
Explanation
1. 특정 deferred transaction id를 지우는 경우
기본적으로 deferred transaction queue에 쌓인 트랜잭션을 지우는
방법은 다음과 같다.
dbms_defer_sys.delete_tran(deferred_tran_id, destination)
즉 다음 예와 같다.
SQL>exec dbms_defer_sys.delete_tran('2.7.10', 'rep2.world');
이때, 해당 destination에 대한 모든 트랜잭션인경우는 앞의 argument를
null로 하고, 특정 transaction id에 대한 모든 destination에 대해서인
경우는 뒷부분의 argument를 null로 한다.
결국 저장된 모든 deferred transaction이라면, 다음과 같다.
SQL>exec dbms_defer_sys.delete_tran(null,null);
2. 특정 table에 관한 내용만 지우는 경우
예를 들어 특정 table, 여기서는 DEPT table에 관한 사항을 DEFCALL에서
지우고자 한다면 다음과 같이 조치하면 된다.
SQL>connect repadmin/repadmin
SQL>set pagesize 1000
SQL>set head off
SQL>spool purgedefcall.sql
SQL>select 'exec dbms_defer_sys.delete_tran('''
|| deferred_tran_id || ''', null);'
from defcall
where packagename like 'DEPT%';
SQL>spool off
spool에 의해 만들어진 purgedefcall.sql을 깨끗하게 편집한 후 다시 save한다.
SQL>connect repadmin/repadmin
SQL>@purgedefcall.sql
이때 만약 특정 site로의 전달만을 막고자 한다면, null대신 MS_B.WORLD와 같이
해당 site를 가리키는 database link이름을 직접 지정하면 된다.
3. 전체 queue의 내용을 모두 지우는 경우
DEFCALL의 내용을 모두 지우는 경우라면 기본적으로는 앞에서 사용한
DBMS_DEFER_SYS.DELETE_TRAN을 이용하면 된다.
SQL>connect repadmin/repadmin
SQL>exec dbms_defer_sys.delete_tran(null,null);
DEFERROR의 내용을 모두 지우는 경우에는 DBMS_DEFER_SYS.DELETE_ERROR를
사용한다.
SQL>exec dbms_defer_sys.delete_error(null,null);
그런데 이 delete_tran과 delete_error의 경우는 내부적으로 delete문장을
사용하면서 undo record를 위해 rollback을 사용하면서 지워야 하는 데이타가
매우 많은 경우 속도도 문제가 되고 rollback space오류도 발생 가능하다.
이러한 경우에는 다음과 같이 truncate command를 이용하여 간단하고 빠르게
deferred transaction queue의 내용을 정리할 수 있다.
(1) Oracle7의 경우
DEF$_CALL, DEF$_CALLDEST, DEF$_ERROR 를 모두 truncate시킨다.
단 이때 DEF$_CALLDEST가 DEF$_CALL을 reference하는 constraint가 있는 관계로,
DEF$_CALLDEST를 모두 truncate하여 데이타가 전혀 없는 상태에서도,
DEF$_CALL이 truncate가 되지 않는다.
delete operation의 경우 child table이 비어 있다면 master table의 데이타를
지우는데 오류가 없지만, truncate의 경우는 데이타 확인 없이 바로 지우는
것이기 때문에 child table에 데이타가 없다하더라도 그러한 check없이,
무조건 자신을 reference하는 child table의 constraint가 enable되어 있는한은
master table이 truncate가 불가능하게 된다.
SQL>connect as system/password
SQL>alter table system.DEF$_CALLDEST disable constraint
DEF$_CALLDEST_CALL;
SQL>truncate table system.DEF$_CALL;
SQL>truncate table system.DEF$_CALLDEST;
SQL>truncate table system.DEF$_ERROR;
SQL>alter table system.DEF$_CALLDEST enable constraint DEF$_CALLDEST_CALL;
(2) Oracle8의 경우
Oracle8에서는 DEF$_CALLDEST가 더이상 DEF$_CALLDEST_CALL constraint를
가지지 않으므로 이 부분에 대한 고려는 필요없이 다음과 같이 해당 table들을
truncate시키면 된다.
SQL>truncate table system.DEF$_AQCALL;
SQL>truncate table system.DEF$_CALLDEST;
SQL>truncate table system.DEF$_ERROR;
SQL>truncate table system.DEF$_AQERROR;
4. deferror의 내용을 지우는 방법
deferred transaction이 remote로 전달되어 반영되다가 오류가 발생하면
source가 되는 데이타베이스의 queue에서는 해당 내용이 사라지고,
반영되던 destination 데이타베이스의 deferror와 defcall/deftran에
해당 내용이 쌓이게 된다.
이러한 경우 error의 내용을 다시 반영을 시도하거나 아니면 내용을
확인 후 지우게 된다.
다음과 같이 지우면 된다.
SQL>exec dbms_defer_sys.delete_error(null,null);
참고로 다시 수행하는 것은
SQL>exec dbms_defer_sys.execute_error(null,null);
Reference Documents
<Note:190885.1> How to Clear Down the Deferred Queue and DBMS_DEFER_SYS.
DELETE_TRAN

Replication/ Master Site

Similar Messages

Maybe you are looking for