Multi-Master Requires Cold Backup?
Since the restore of a multi-master instance requires the restore of the backup and the changelog directories, and since the changelog cannot be backuped up with anything other than an OS copy command, and since an OS copy command will usually not create a valid backup of an open database file, is it not true that multi-master backups must therefore be cold backups?
By "In flight changes" I mean changes which have occurred on one master replica, but have not been replicated to the other. My mindset is on multi-master, but I suppose much of this applies to any supplier / consumer backup/restore (not just multi-master).
Your cold backup procedure is what I have implemented until 5.2. I still have not had time to test restores fully at this point with this backup procedure. I am somewhat concerned that after restoring both instances from their own backups, will replication work and will they synch up correctly.
Similar Messages
-
Oracle version : 11G release 2
Platform : AIX
After going through various OTN posts, i wrote the below Cold backup script (Untested). If there are any unnecessary stuff or enhancements required, please let me know
$ rman target /
RMAN>
run
allocate channel c1 type disk format '/u05/rmanbkp/%d_COLD_DB_%u';
sql 'alter system archive log current';
shutdown immediate;
startup mount;
backup database including current controlfile tag='mydbname_full_bkp';
backup spfile tag = 'mydbname_SPFILE';
release channel c1;
}Hi T.Boyd
What do you guys think of the backup script? Any room for improvement? I found in a small percentage of the shutdown immediate commands issued, the instance hangs (specially when the machine is very busy).
I have modified my rman coldbackup the procedure to:
shutdown abort;
host 'sleep 3';
startup restrict;
shutdown immediate;
startup mount;Maybe you can add more channels to improve performance. You can use more resources as there are no users on the database anyway....
Regards,
Tycho -
A question about restoring from cold backup(control file backup not clear)
Hi,
I had another question about restoring the cold backup. My database is in noarchivelog mode and after taking a consistent cold backup, all I need to do is to restore the backup right? -Why I got this question is because: when I backup my control file to trace, I see statements like this:-----
-- Commands to re-create incarnation table
-- Below log names MUST be changed to existing filenames on
-- disk. Any one log file from each branch can be used to
-- re-create incarnation records.
-- ALTER DATABASE REGISTER LOGFILE '/uo1/app1/arch1_1_647102958.dbf';
-- Recovery is required if any of the datafiles are restored backups,
-- or if the last shutdown was not normal or immediate.
RECOVER DATABASE
-- Database can now be opened normally.
ALTER DATABASE OPEN;
My database is in noarchivelog mode now so don't know why these statements (of register the logfile) is there in the backup of control file? so when I restore the cold backup of this database, it will still work correct? (there is no logfile I have only CRD files in cold backup -no archive log files.)
thanks
NiravThanks for your inputs! It is most useful to me.
Regards
Nirav -
Hi All;
OS: Windows
DB: 10.2.0.3
I am trying to schedule a cold backup, I was wondering if I can just down the oracle database service and then directly copy the files using copy command and post the completion of copy of the files I will up the oracle services.
Will this method work I cannot test the same as I do not have access to the server right now.
Regards,
SphinxYes you can, by default stopping the service will execute a SHUTDOWN IMMEDIATE. To check this, look at the registry keys named something like
HKEY_LOCAL_MACHINE\SOFTWARE\Oracle\KEY_OraDb12c_home1\ORA_ORCL_SHUTDOWN
HKEY_LOCAL_MACHINE\SOFTWARE\Oracle\KEY_OraDb12c_home1\ORA_ORCL_SHUTDOWNTYPE
and you'll see the settings.
John Watson
Oracle Certified Master DBA
http://skillbuilders.com -
Cold backup tablespace restore
From a cold backup can you restore a tablespace to a different database.
A datafile was created and dropped and now we are recieving ORA-1186, ORA-1157 cannnot/identify lock datafile
We know that if our database goes down it won't come back up. The table space has about 250 tables and its huge about 100g in size. Does anyone know of what steps needs to be taken?
Edited by: user10767182 on Jan 6, 2009 8:35 PMI couldn't quite workout what you were saying, but I think you were suggesting that copies of the lost tables and data are sitting in a second database somewhere, and you would like to pull them out of that database and plug them into the broken database. Is that right?
If so, you cannot take a datafile from one database and plug it into another, unless you use the transportable tablespace option.
Basically, on your broken database, you'd shut it down, bring it back to the mount state and then say alter database datafile X offline drop
That will let you issue an alter database open followed by a drop tablespace X, and your broken database will at least be open, minus the important tablespace
You then get your second database open and make the important tablespace read-only
You'd drop to the command line and do an export using the TRANSPORT_TABLESPACE option -the command is too susceptible to the specifics to show you here. Check the documentation at http://download.oracle.com/docs/cd/B19306_01/backup.102/b14194/rcmsynta063.htm
You then copy the datafile and the export dump file to the server where your broken database is running
You then run an import, again specifying the TRANSPORT_TABLESPACE option
Effectively, the datafile copy gets 'plugged in' to the broken database and gets adopted as a native, brand new tablespace, complete with contents. You finish off by making the tablespace read-write in both databases once more.
Obviously, you lose data using this sort of process: the data comes back into your 'broken' database in the same state it was in your second database, and you can't apply redo to it to recover it to a more recent state. But that's going to be the best you can do if you don't have proper physical backups of the file. Your subject mentioning cold backups confused me a little on that score too.
So I won't go into any more detail for now. It may be that I misunderstood the reference to 'restore a tablespace to a different database' and your requirements completely. But if this sounds like what you are after, and if you are stuck on any of the details, then you can always post a follow-up. -
Hello All ,
I am new to the DBA activities..,kindly help me with the cold backup cloning.
We have a backup of PROD(cold bkp) taken 2 weeks before.
After the backup was taken we have dropped a tablespace(with 4 dbfs) as per the customer requirement.
And also we have created a new tablespace (with 1 dbf).
So the cold backup does not include the new dbfs created which is needed,and it does include the dropped dbfs which are not needed.
Now we are going to do a clone on to the TEST with the above cold backup,Could anybody please help me how to proceed..?
Thanks in advance...
Thanks,
Chan.My thoughts were :
If the test and prod hosts are identical:
1. restore the cold backup to test.
2. Copy over all archive logs created since to the archive log destination.
3. Use a new cold backup of current production controlfiles to overwrite those restored to test, or if prod has to stay up , use RMAN to create a backup of the controlfile & spfile from production that you then restore using rman on test (with DB started nomount & DBID set).
4. Open test 'mount' , recover database, then open resetlogs.
Or am I forgetting something?
I vaguely remember that there's something about changes to DB structure (additional Tablespaces and files) and control files thus other posts refering to 'alter database create file' ...
Regards,
Vin.
PS. You might just be better of creating a new Rman compressed backupset database backup and restoring that. -
Cloning Using Cold Backups !!!!
Dear All,
Greetings,
I Want To Ask What Are All The Files And Directories That I Need To Copy From The Original Server To The Destination To Perform A Cloning From Cold Backups ?
The Second Thing, Can You Plz, Tell Me The Complete Steps To Perform This Cloning .
Plz, It's Urgent
Thank You .
GuivaraHi Guivara,
Sabar given urls that would be useful for your requirement.
Generally for cold backups
We need to take the below's from your original server.
Control Files,
Redo log files,
all Data files
After restore those files into your destination, you should edit/modify the original server files to destination server path.
For example,
System.dbf is located on /u01/oracle/product/9iR2/oradata/' but in destination server you want to put other location, then you should change the path using
Alter database rename .....
Cheers,
Moorthy.GS -
Rebuild from cold backup on windows
Greetings,
Our Db was destroyed. We have a new machine now, and I have loaded the OS ( windows NT ), and the Db software ( Oracle 9i ). I have a cold backup of the old database that I would like to use to get the old system back. Where can I find documentation explaing the required steps to getting the Database back on line ( on Windows as a service )using a cold backup?
Any help or advice are greatly appreciated.
Thank You.Hello,
If the location of the files didn't change reconstructing the database instance from coldbackup takes just few steps
1.) Create an instance
Oradim ânew âsid testdb -mode a
2.) Connect to the instance
Sqlplus /nolog
Connect / as sysdba Ã
3.) Start the database
Startup pfile=localtion to inti.ora
Let us know your case details and I would perhaps be able to help. I experienced the same problem and successfully rebuilt the database
Ref:
I have modified some Steps in manually creating a database to suit to my needs
http://download-west.oracle.com/docs/cd/A91202_01/901_doc/server.901/a90117/create.htm#1019233
Clone db: http://www.dbaclick.com/forums/archive/16/35.html
Hope it helps!! -
Hi,
Our client has provided us with the cold Backup (the file system) of the pre-prod system which runs Oracle 9.2.0.8.0 on HP-Unix.
We want to restore this backup on Windows machine.
Can anybody provide help or some document which can guide us through the steps to setup new instance to restore the cold backup ?
Is this the same as creating standby system ?
Thanks a lot :)>> we are trying to re-create the database to use the available backup. I think this is a somewhat like creating a standby database (without continuous synchronization)
In that case, you can not utilize your cold backup on Windows machine at all, Until and unless you have the same OS, you can not prepare the standby database.
Please read the below URL for requirements for deploying Standby.
http://download.oracle.com/docs/cd/B10501_01/server.920/a96653/considerations.htm#52221
Regards,
Sabdar Syed. -
제품 : ORACLE SERVER
작성날짜 : 2003-08-04
Automating cold backup on NT
============================
unix에서는 cron 이라는 것을 이용해서 주기적으로 데이타베이스를 원하는 시점에
shutdown 시킨 후 cold backup 을 할수 있다.
그렇다면 nt에서는 어떻게 이런 작업을 수행해 줄 수 있을까.
여기서는 이렇게 주기적으로 database 를 cold backup받는 방법에 관해서 설명
하고자 한다.
데이타베이스 환경은 다음과 같이 가정한다.
ORACLE_HOME = c:\ORANT
ORACLE_SID = ORCL
데이타화일들은 c:\ORANT\DATABASE 디렉토리에 존재한다.
cold backup을 schedule하기 위해서는 windows NT의 AT scheduler utility 가
필요하며
이것의 GUI version은 windows NT의 Resource Kit에 포함되어 있다.
우선은 NT scheduler service가 동작하고 있는지 확인해야 한다.
확인 방법은 NT 제어판을 열고 Service icon을 더블 클릭하여 schedule service가
동작하고 있으며 automatic인지 확인한다.
그러면 schedule 방법을 다음과 같이 정리하였다.
1. 만약 command line scheduler를 사용하고 있다면 다음과 같은 작업을 해준다.
backup.bat라는 batch file 을 만들어서 다음과 같은 내용을 입력한다.
call c:\orant\bin\svrmgr23.exe command="@c:\shutdown.sql"
copy c:\orant\database\*orcl.ora c:\backup
call c:\orant\bin\svrmgr23.exe command="@c:\startup.sql"
backup을 schedule 한다.
NT의 command prompt(Dos prompt)에서 다음과 같은 형식으로 입력한다.
AT time "C:\backup.bat"
eg. C:\> at 14:45 "c:\backup.bat"
2. SHUTDOWN.SQL SCRIPT를 다음과 같이 만들어 준다.
connect internal/password
shutdown immediate
exit
3. STARTUP.SQL SCRIPT를 다음과 같이 만들어 준다.
connect internal/password
startup pfile=C:\ORANT\DATABASE\INITORCL.ORA
exit
이상과 같이 해주면 원하는 시간에 db를 cold backup 받을 수 있을 것이다.
Reference Documents
<Note:74981.1>You must cleanly shutdown database before proceeding with a cold consistent backup, otherwise you compromise database integrity, so recoverability.
By clean shutdown I mean Normal, Transactional or Immediate, never abort. The point is that the last transaction must properly be recorded by the last checkpoint, and no recovery process is required. Additionally, a clean shutdown guarantees consistency acrorss all datafiles.
~ Madrid. -
Partition Restore Cold Backup Different System and Different Version DB
Is it possible, can we restore the cold backup at different database version and different operating system(but same ENDIAN_FORMAT) ????????
My original system HP-UX (64-bit) and database version 10.2.0.3.0 i have cold backup and i want to partion restore on Linux Redhot 5.5 (64-bit), database version 10.1.0.4.0 .
If it is possible. How can i do this.My original system HP-UX (64-bit) and database version 10.2.0.3.0 i have cold backup and i want to partion restore on Linux Redhot 5.5 (64-bit), database version 10.1.0.4.0 .1. AFAIK you can not do it i mean wiht cold backup move from HP to linux
2. You can try to use exp-imp if you want to change db version.
See below notes:
Master Note For Oracle Database Upgrades and Migrations [ID 1152016.1]
Different Upgrade Methods For Upgrading Your Database [ID 419550.1]
Regard
Helios -
Two problems after reinitializing multi-master replication.
Hi,
I recently set up multi-master replication between two machines, A and B, running iDS 5.1. Everything works well until I disable replication and re-enable replication. Here's the procedure that I follow:
1. Delete agreement on machine A and B.
2. Disable the replica on machine A and B.
3. Disable the changelog on machine A and B.
4. Enable the changelog on machine A and B.
5. Enable the replica on machine A and B.
6. Create replication agreement on A and B.
7. Initialize B from A.
Problem #1:
After I follow this procedure, updates to A result in the error:
<B>
Data required to update replica "cn=rep1, cn=replica, cn="o=sylantro.com", cn=mapping tree, cn=config (host rosemary.sylantro.com, port 389)" has been purged. The replica must be reinitialized.
[07/Feb/2002:10:43:28 -0800] NSMMReplicationPlugin - Replication: Incremental update failed and requires administrator action
</B>
At this point if I re-initialize B from A, the problem ceases and everything is fine.
This problem still occurs even if I restart the server before Step 4.
Problem #2.
When I assign machine A the old replicaId of machine B and assign machine B the old replicaID of machine A in Step 5, I get the following error.
<B>
[07/Feb/2002:10:06:02 -0800] NSMMReplicationPlugin - Unable to aquire replica "cn=rep1, cn=replica, cn="o=s
ylantro.com", cn=mapping tree, cn=config (host wasrocky.sylantro.com, port 389)": the replica has the same Replica ID as this one. Replication is aborting.
</B>
The server looks like it is remembering its old replicaID.
This problem still occurs even if I restart the server before Step 4.
Any insight would be greatly appreciated.
Thanks!
GilFor the problem #1, the error message is normal since you have actually remove the changelog. However, once you have done an initialization, Replication should work.
For the problem #2, the message is correct, because, although you have changed the ReplicaId on both server, your data still contain references to the previous replicaId both on Server A and Server B.
In general, Replica ID uniquely identify a Master in a replication topology. It should not change. The replicaID is also used in the timestamps associated with each change and they guarantee that the Replication can always distinguish 2 changes, and also can detect time synchronization issues between masters.
One way to avoid the Replica ID issue when changing ReplicaIds, is to reload the data on one server and then re-initialize the second one from the first one.
Regards,
Ludovic. -
DS 5 SP1 multi-master replication
I am running DS5 SP1 on Solaris 8 with the recommended patch set. I am
using replica id 1 for master1 and replica id 2 for master2.
I am replicating 7 nodes in a multi-master scenario. I start with about
1000 base entries in each db node, and add approximately 1000 to each
node via a meta directory interface to master1. I run 9 passes, so I am
adding approximately 10,000 entries to each node over the course of the
passes.
Occasionally, inconsistently, and without explanation, slapd stops on my
master1 (sometimes master2). I have successfully run through all passes
at differing times, and then it happens. There is no specific error
message indicating it has died, but a ps reveals that it is no longer
running. Once I restart it, the error log indicates that the directory
had a disorderly shutdown. It has the nasty effect of resetting my
changelogs to 0 and requiring me to reinitialize my replication
agreements. Below is a relevant excerpt from my error log, hopefully
someone can give me some insight as to why this is happening . . . . .
[10/Dec/2001:17:13:55 +0000] NSMMReplicationPlugin - Failed to connect
to replication consumer master2.domain.com:636
[10/Dec/2001:17:13:57 +0000] NSMMReplicationPlugin - Could not send
consumer master2.domain.com:636 the bind request
[10/Dec/2001:17:13:57 +0000] NSMMReplicationPlugin - Failed to connect
to replication consumer master2.domain.com:636
[10/Dec/2001:17:14:02 +0000] NSMMReplicationPlugin - Could not send
consumer master2.domain.com:636 the bind request
[10/Dec/2001:17:14:02 +0000] NSMMReplicationPlugin - Failed to connect
to replication consumer master2.domain.com:636
[10/Dec/2001:17:14:08 +0000] NSMMReplicationPlugin - Could not send
consumer master2.domain.com:636 the bind request
[10/Dec/2001:17:14:08 +0000] NSMMReplicationPlugin - Failed to connect
to replication consumer master2.domain.com:636
[10/Dec/2001:20:29:03 +0000] NSMMReplicationPlugin - Could not send
consumer master2.domain.com:636 the bind request
[10/Dec/2001:20:29:03 +0000] NSMMReplicationPlugin - Failed to connect
to replication consumer master2.domain.com:636
[10/Dec/2001:20:29:03 +0000] NSMMReplicationPlugin - received error 81:
NULL from replica master2.domain.com:636 for add operation
[10/Dec/2001:20:29:03 +0000] NSMMReplicationPlugin - Failed to replay
change (uniqueid 45e39dd9-1dd211b2-80a8c58f-66391b25, CSN
3c151ad00001000
10000) to replica "cn=to master2, cn=replica,
cn="ou=group,o=company,c=us", cn=mapping tree, cn=config (host
master2.domain.com, port 636)"
: Local error. Will retry later.
[10/Dec/2001:20:29:03 +0000] NSMMReplicationPlugin - failed to send
extended operation to replica master2.domain.com:636; LDAP error - 81
[10/Dec/2001:20:29:03 +0000] NSMMReplicationPlugin - Warning: unable to
send endReplication extended operation to consumer "cn=to master2,
cn=replica, cn="ou=group,o=company,c=us", cn=mapping tree, cn=config
(host master2.domain.com, port 636)" - error 1
[10/Dec/2001:20:29:13 +0000] NSMMReplicationPlugin - failed to send
extended operation to replica master2.domain.com:636; LDAP error - 81
[10/Dec/2001:20:29:13 +0000] NSMMReplicationPlugin - Unable to send a
startReplication extended operation to consumer "cn=to master2,
cn=replica, cn="ou=group,o=company,c=us", cn=mapping tree, cn=config
(host master2.domain.com, port 636)". Will retry later.
[10/Dec/2001:20:29:13 +0000] NSMMReplicationPlugin - Could not send
consumer master2.domain.com:636 the bind request
[10/Dec/2001:20:29:13 +0000] NSMMReplicationPlugin - Failed to connect
to replication consumer master2.domain.com:636
[10/Dec/2001:20:29:17 +0000] NSMMReplicationPlugin - Could not send
consumer master2.domain.com:636 the bind request
[10/Dec/2001:20:29:17 +0000] NSMMReplicationPlugin - Failed to connect
to replication consumer master2.domain.com:636
[10/Dec/2001:20:29:33 +0000] NSMMReplicationPlugin - failed to send
extended operation to replica master2.domain.com:636; LDAP error - 81
[10/Dec/2001:20:29:33 +0000] NSMMReplicationPlugin - Unable to send a
startReplication extended operation to consumer "cn=to master2,
cn=replica, cn="ou=group,o=company,c=us", cn=mapping tree, cn=config
(host master2.domain.com, port 636)". Will retry later.
[10/Dec/2001:20:30:13 +0000] NSMMReplicationPlugin - failed to send
extended operation to replica master2.domain.com:636; LDAP error - 81
[10/Dec/2001:20:30:13 +0000] NSMMReplicationPlugin - Unable to send a
startReplication extended operation to consumer "cn=to master2,
cn=replica, cn="ou=group,o=company,c=us", cn=mapping tree, cn=config
(host master2.domain.com, port 636)". Will retry later.
[10/Dec/2001:20:31:54 +0000] - iPlanet-Directory/5.0 ServicePack 1
B2001.264.1425 starting up
[10/Dec/2001:20:31:56 +0000] - Detected Disorderly Shutdown last time
Directory Server was running, recovering database.
[10/Dec/2001:20:34:38 +0000] NSMMReplicationPlugin - replica_reload_ruv:
Warning: data for replica ou=node3,ou=group,o=company,c=us was reloaded
and it no longer matches the data in the changelog. Recreating the
changlog file. This could effect replication with replica's consumers in
which case the consumers should be reinitialized.
[10/Dec/2001:20:34:38 +0000] NSMMReplicationPlugin - replica_reload_ruv:
Warning: data for replica ou=node1,ou=group,o=company,c=us
was reloaded and it no longer matches the data in the changelog.
Recreating the changlog file. This could effect replication with
replica's consumers in which case the consumers should be reinitialized.
[10/Dec/2001:20:34:38 +0000] NSMMReplicationPlugin - replica_reload_ruv:
Warning: data for replica ou=node4,ou=group,o=company,c=us was reloaded
and it no longer matches the data in the changelog. Recreating the
changlog file. This could effect replication with replica's consumers in
which case the consumers should be reinitialized.
[10/Dec/2001:20:34:38 +0000] NSMMReplicationPlugin - replica_reload_ruv:
Warning: data for replica ou=group,o=company,c=us was reloaded a
nd it no longer matches the data in the changelog. Recreating the
changlog file. This could effect replication with replica's consumers in
which case the consumers should be reinitialized.
[10/Dec/2001:20:34:38 +0000] NSMMReplicationPlugin - replica_reload_ruv:
Warning: data for replica ou=node2,ou=group,o=company,c=us was reloaded
and it no longer matches the data in the changelog. Recreating the
changlog file. This could effect replication with replica's consumers in
which case the consumers should be reinitialized.
[10/Dec/2001:20:34:38 +0000] NSMMReplicationPlugin - replica_reload_ruv:
Warning: data for replica ou=node5,ou=group,o=company,c=us was reloaded
and it no longer matches the data in the changelog. Recreating the
changlog file. This could effect replication with replica's consumers in
which case the consumers should be reinitialized.
[10/Dec/2001:20:34:38 +0000] NSMMReplicationPlugin - replica_reload_ruv:
Warning: data for replica ou=node6,ou=group,o=company,c=us was reloaded
and it no longer matches the data in the changelog. Recreating the
changlog file. This could effect replication with replica's consumers in
which case the consumers should be reinitialized.
[10/Dec/2001:20:34:39 +0000] - slapd started. Listening on all
interfaces port 389 for LDAP requests
[10/Dec/2001:20:34:39 +0000] - Listening on all interfaces port 636 for
LDAPS requests
[10/Dec/2001:20:34:39 +0000] - cos_cache_getref: no cos cache created
[10/Dec/2001:20:34:44 +0000] NSMMReplicationPlugin - Data required to
update replica "cn=to master2, cn=replica,
cn="ou=node1,ou=group,o=company,c=us", cn=mapping tree, cn=config (host
master2.domain.com, port 636)" has been purged. The replica must be
reinitialized.
[10/Dec/2001:20:34:44 +0000] NSMMReplicationPlugin - Replication:
Incremental update failed and requires administrator action
[10/Dec/2001:20:35:02 +0000] NSMMReplicationPlugin - Data required to
update replica "cn=to master2, cn=replica, cn="ou=node2,ou=group,o=
company,c=us", cn=mapping tree, cn=config (host master2.domain.com, port
636)" has been purged. The replica must be reinitialized.
[10/Dec/2001:20:35:03 +0000] NSMMReplicationPlugin - Replication:
Incremental update failed and requires administrator action
[10/Dec/2001:20:36:50 +0000] NSMMReplicationPlugin - Data required to
update replica "cn=to master2, cn=replica,
cn="ou=node3,ou=group,o=company,c=us", cn=mapping tree, cn=config (host
master2.domain.com, port 636)" has been purged. The replica must be
reinitialized.
[10/Dec/2001:20:36:51 +0000] NSMMReplicationPlugin - Replication:
Incremental update failed and requires administrator action
[10/Dec/2001:20:38:34 +0000] NSMMReplicationPlugin - Data required to
update replica "cn=to master2, cn=replica, cn="ou=node4,ou=group,o=
company,c=us", cn=mapping tree, cn=config (host master2.domain.com, port
636)" has been purged. The replica must be reinitialized.
[10/Dec/2001:20:38:34 +0000] NSMMReplicationPlugin - Replication:
Incremental update failed and requires administrator action
[10/Dec/2001:20:42:31 +0000] NSMMReplicationPlugin - Data required to
update replica "cn=to master2, cn=replica, cn="ou=node5,ou=group,o=
company,c=us", cn=mapping tree, cn=config (host master2.domain.com, port
636)" has been purged. The replica must be reinitialized.
[10/Dec/2001:20:42:31 +0000] NSMMReplicationPlugin - Replication:
Incremental update failed and requires administrator action
[10/Dec/2001:20:44:18 +0000] NSMMReplicationPlugin - Data required to
update replica "cn=to master2, cn=replica,
cn="ou=node6,ou=group,o=company,c=us", cn=mapping tree, cn=config (host
master2.domain.com, port 636)" has been purged. The replica must be
reinitialized.
[10/Dec/2001:20:44:19 +0000] NSMMReplicationPlugin - Replication:
Incremental update failed and requires administrator action
[10/Dec/2001:20:46:06 +0000] NSMMReplicationPlugin - Data required to
update replica "cn=to master2, cn=replica, cn="ou=group,o=company,c=us",
cn=mapping tree, cn=config (host master2.domain.com, port 636)" has been
purged. The replica must be reinitialized.
[10/Dec/2001:20:46:06 +0000] NSMMReplicationPlugin - Replication:
Incremental update failed and requires administrator actionHi Jennifer,
There's nothing in the logs that explain why your server is disappearing.
I think that the reseting of the changelog is a known problem that should be fixed in 5.0SP2 and is fixed in iDS 5.1 which is now available.
Just one quick question, when you say that you have 7 nodes, you mean 7 container entries or 7 suffixes that are separated replication areas ?
Regards,
Ludovic. -
Streams multi master replication problem
Dear Stefan Menschel,
Good morning
DB:10.2.0.2
OS:windowsXp
This is we are trying in test environment
I have downloaded the OSC and just configured the multi master replication i mean just run the folder(400) scripts. After all the scripts exec. i was monitoring the streams activity through OEM console.
i have got an error in -->streams -->propagation-->topology
ERROR : ORA-25205: the QUEUE S1ADMIN.A_QN1_DBS2TEST does not exist
and link is showing as RED culler with some error.
hub : DBS1TEST
streamsadmin: s1admin
target1 : DBS2TEST
streamsadmin: s2admin
i might be wrong while preparing the scripts. if please let me know.
i can send you the scripts also
Thanks
Hareesh.L
DBA mailid: [email protected]
Message was edited by:
user627795I've never seen that error, but do make sure that your Replica ID's are different (that's a requirement). Also, you might consider setting up one of the masters as a hub at first, then initialize it, and then promote it. Once you've done that, you can create the replication agreement from the second host to the first one.
Patrick -
DSCC multi-master replication issue
Hello All,
I am trying to setup 2 DSCC consoles with multi-master replication enabled(cn=dscc), facing issue when I see directory server list in both dscc consoles, I see below 2 dscc instances also which should not be there(since they are ADS, should be hidden). Also the changes does not reflect immediately, takes around 30 minutes or so.
Please note I am running 2 ADS instances on one box with port no 3998 and 4000 both are master. Seek your guidance on how to fix this issue.
localhost:3998 (server not registered) - Started -
localhost:4000 (server not registered) - Started -
Below are the steps I carried out to setup multi-master replication-
On instance 1
Check the DSCC port no of instance 1
D:\ldap_server\ds6\bin>dsadm info d:\ldap_server\var\dscc6\dcc\ads
Instance Path: d:/ldap_server/var/dscc6/dcc/ads
Owner: AT0094060
Non-secure port: 3998
Secure port: 3999
Bit format: 32-bit
State: Running
Server PID: 2820
DSCC url: -
Windows service registration: Disabled
Instance version: D-A00
Enable replication-
D:\ldap_server\ds6\bin>dsconf enable-repl -h localhost -p 3998 -e -d 10 master cn=dscc
Enter "cn=Directory Manager" password:
Use "dsconf create-repl-agmt" to create replication agreements on "cn=dscc".
Setup repl agmt
D:\ldap_server\ds6\bin>dsconf create-repl-agmt -h localhost -p 3998 -e cn=dscc localhost:4000
Enter "cn=Directory Manager" password:
Use "dsconf init-repl-dest cn=dscc localhost:3998" to start replication of "cn=dscc" data.
Setup rep password
D:\ldap_server\ds6\bin>dsconf set-server-prop -h localhost -p 3998 -D "cn=directory manager" -e def-repl-manager-pwd-file:d:\rmpassword.txt
Enter "cn=Directory Manager" password:
Check the password
D:\ldap2_server\ds6\bin>dsconf get-server-prop -h localhost -p 3998 -e def-repl-manager-pwd
Enter "cn=Directory Manager" password:
def-repl-manager-pwd : {SSHA}g9OpeO2H57MH2Eq4xV5gbxVqHGzEG2VpdBSuIA==
Restart ADS to read new changes
D:\ldap_server\ds6\bin>dsadm restart d:\ldap-server\var\dscc\dcc\ads
Check suffix prop-
D:\ldap_server\ds6\bin>dsconf get-suffix-prop -h localhost -p 3998 -e cn=dscc
Enter "cn=Directory Manager" password:
all-ids-threshold : inherited (4000)
db-name : bellatonus
db-path : D:/ldap_server/var/dscc6/dcc/ads/db/bellatonus
enabled : on
entry-cache-count : unlimited
entry-cache-size : 10M
entry-count : 12
moddn-enabled : inherited (off)
parent-suffix-dn : undefined
referral-mode : disabled
referral-url : ldap://machine1:4000/cn%3Ddscc
repl-accept-client-update-enabled : on
repl-cl-max-age : 1w
repl-cl-max-entry-count : 0
repl-id : 10
repl-manager-bind-dn : cn=replication manager,cn=replication,cn=config
repl-purge-delay : 1w
repl-rewrite-referrals-enabled : off
repl-role : master
require-index-enabled : off
Run accord-
D:\ldap_server\ds6\bin>dsconf accord-repl-agmt -h localhost -p 3998 -e cn=dscc localhost:4000
To test replication manager password use-
ldapsearch -h localhost -p 3998 -D "cn=replication manager,cn=replication,cn=config" -q -b "" -s base objectclass=*namingContexts
Please enter bind password:
check the replication status
D:\ldap2_server\ds6\bin>dsconf show-repl-agmt-status -h localhost -p 3998 -e cn=dscc localhost:4000
Enter "cn=Directory Manager" password:
Configuration Status : OK
Authentication Status : OK
Initialization Status : OK
Status : Enabled
Last Update Date : Jun 13, 2012 4:04:22 PM
On instance 2
Check the DSCC port no-
D:\ldap_server\ds6\bin>dsadm info d:\ldap2_server\var\dscc6\dcc\ads
Instance Path: d:/ldap2_server/var/dscc6/dcc/ads
Owner: AT0094060
Non-secure port: 4000
Secure port: 4001
Bit format: 32-bit
State: Running
Server PID: 4264
DSCC url: -
Windows service registration: Disabled
Instance version: D-A00
Enable replication
D:\ldap_server\ds6\bin>dsconf enable-repl -h localhost -p 4000 -e -d 10 master cn=dscc
Enter "cn=Directory Manager" password:
Use "dsconf create-repl-agmt" to create replication agreements
on "cn=dscc".
Setup repl agmt
D:\ldap_server\ds6\bin>dsconf create-repl-agmt -h localhost -p 4000 -e cn=dscc localhost:3998
Enter "cn=Directory Manager" password:
Use "dsconf init-repl-dest cn=dscc localhost:3998" to start replication of "cn=dscc" data.
Setup repl password
D:\ldap_server\ds6\bin>dsconf set-server-prop -h localhost -p 4000 -D "cn=directory manager" -e def-repl-manager-pwd-file:d:\rmpassword.txt
Enter "cn=Directory Manager" password:
Check the password
D:\ldap2_server\ds6\bin>dsconf get-server-prop -h localhost -p 4000 -e def-repl-manager-pwd
Enter "cn=Directory Manager" password:
def-repl-manager-pwd : {SSHA}g9OpeO2H57MH2Eq4xV5gbxVqHGzEG2VpdBSuIA==
Restart ADS
D:\ldap_server\ds6\bin>dsadm restart d:\ldap2-server\var\dscc\dcc\ads
test replication manager password with
ldapsearch -h localhost -p 4000 -D "cn=replication manager,cn=replication,cn=config" -q -b "" -s base objectclass=*namingContexts
Please enter bind password:
D:\ldap2_server\ds6\bin>dsconf get-suffix-prop -h localhost -p 4000 -e cn=dscc
Enter "cn=Directory Manager" password:
all-ids-threshold : inherited (4000)
db-name : bellatonus
db-path : D:/ldap2_server/var/dscc6/dcc/ads/db/bellatonus
enabled : on
entry-cache-count : unlimited
entry-cache-size : 10M
entry-count : 12
moddn-enabled : inherited (off)
parent-suffix-dn : undefined
referral-mode : disabled
referral-url : ldap://machine1:3998/cn%3Ddscc
repl-accept-client-update-enabled : on
repl-cl-max-age : 1w
repl-cl-max-entry-count : 0
repl-id : 20
repl-manager-bind-dn : cn=replication manager,cn=replication,cn=config
repl-purge-delay : 1w
repl-rewrite-referrals-enabled : off
repl-role : master
require-index-enabled : off
Initialize ADS2 from ADS1 using the replication agreement:
dsconf init-repl-dest -e -i -h localhost -p 3998 cn=dscc localhost:4000
Delete
Check the replication status
D:\ldap2_server\ds6\bin>dsconf show-repl-agmt-status -h localhost -p 4000 -e cn=dscc localhost:3998
Enter "cn=Directory Manager" password:
Configuration Status : OK
Authentication Status : OK
Initialization Status : OK
Status : Enabled
Last Update Date : Jun 13, 2012 4:07:36 PM
Run insync
D:\ldap2_server\ds6\bin>insync -D "cn=directory manager" -j d:\dmpw.txt -s localhost:3998 -c localhost:4000 20
ReplicaDn Consumer Supplier Delay
cn=dscc localhost:4000 localhost:3998 0
cn=dscc localhost:4000 localhost:3998 0
cn=dscc localhost:4000 localhost:3998 0
^C
D:\ldap_server\ds6\bin>insync -D "cn=directory manager" -j d:\dmpw.txt -s localhost:4000 -c localhost:3998 20
ReplicaDn Consumer Supplier Delay
cn=dscc localhost:3998 localhost:4000 0
cn=dscc localhost:3998 localhost:4000 0
cn=dscc localhost:3998 localhost:4000 0Replicating the ADS instance, ie cn=dscc is not supported and not supposed to work so what you are trying to do is futile.
Maybe you are looking for
-
How can I know what the font(s) is/are used in PDF using 'Preview.app'?
As the subject, thanks in advance!
-
Boot camp with xp can't go into stand by
I was trying to put the computer into standby and it entered the "prepare to standby screen" and just froze. Any idea how to fix this? Thanks
-
How to start apache which embeded in oracle
hi, i run the following command under the root priviledge: $ORACLE_HOME/Apache/Apache/bin/apachectl start. but, the server said: ./apachectl start: httpd could not be started what is the reason? when i first install the o8i, the web server is alive(
-
How can i use net start for iws
Dear all : Is there any body know how to restart one of the instance by command prompt ??? Many Thanks !! Thomas
-
xsetroot: unable to open display '' xset: unable to open display "" xset: unable to open display "" xset: unable to open display "" xsetroot: unable to open display '' startkde: Starting up... startkde: Running kpersonalizer... kwin: cannot conn