Prardhan N
asked on
need to copy the data of a set of tables from instance1 to instance2
Hi All
Greetings!!!
I have 2 instances ( db2inst1 and db2inst2) on the same machine. Each instance is having one DB each. I need to copy the data of a set of tables (for example 20 tables) from
instance1 (db2inst1) to instance2 (db2inst2). This frequency of data copy should be daily once. Every time only Delta data need to be copied from source to target.
What are the possible ways of achieving this.
Greetings!!!
I have 2 instances ( db2inst1 and db2inst2) on the same machine. Each instance is having one DB each. I need to copy the data of a set of tables (for example 20 tables) from
instance1 (db2inst1) to instance2 (db2inst2). This frequency of data copy should be daily once. Every time only Delta data need to be copied from source to target.
What are the possible ways of achieving this.
Hi,
Federation between DB2 LUW databases is available in all editions in DB2 v11.1
If you have Workgroup server or above you could use SQL Replication or CDC replication methods to do this.
https://www.ibm.com/support/knowledgecenter/en/SSTRGZ_10.2.1/com.ibm.swg.im.iis.db.repl.sqlrepl.doc/topics/iiyrscncsqlreplovu.html
Regards,
Tomas Helgi
Federation between DB2 LUW databases is available in all editions in DB2 v11.1
If you have Workgroup server or above you could use SQL Replication or CDC replication methods to do this.
https://www.ibm.com/support/knowledgecenter/en/SSTRGZ_10.2.1/com.ibm.swg.im.iis.db.repl.sqlrepl.doc/topics/iiyrscncsqlreplovu.html
Regards,
Tomas Helgi
ASKER
For a OLTP environment, which approach is suitable either Federation or SQL Replication ?
I have to think from Performance point of view as well. Any inputs with respect to performance, which method is good?
i am using this version -- 10.5.0.7
I have to think from Performance point of view as well. Any inputs with respect to performance, which method is good?
i am using this version -- 10.5.0.7
I had no idea that Federation was generally available starting at release 11.1. Thanks, Tom!!!!
10.5.0.7 is older than 11.1, so Federation isn't an option unless you've got a paid up license.
Still need to know how you're going to determine which rows to move to the second instance.
Kent
10.5.0.7 is older than 11.1, so Federation isn't an option unless you've got a paid up license.
Still need to know how you're going to determine which rows to move to the second instance.
Kent
Hi,
>For a OLTP environment, which approach is suitable either Federation or SQL Replication ?
It highly depends on your overall workload and how your system works.
CDC and SQL Replication works in such way that every data processed in an UOW on System A is transferred from A to B immediately after a commit is issued. You only need to define the tables involved in each subscription (where each subscription monitors and handles a set of tables that takes part in a UOW).
On the other hand, then Federated connection is only suitable for batch-like processing (scheduled and/or event driven) of data between A and B.
You need to make sure that the workload (by looking at how your system works) on instance B (the target) handles the chosen solution of data transfer in a correct manner.
That is. Ask yourself these questions:
Does system B need data immediately after UOW ends on A ?
Does system B need data on a scheduled/event driven basis ?
Regards,
Tomas Helgi
>For a OLTP environment, which approach is suitable either Federation or SQL Replication ?
It highly depends on your overall workload and how your system works.
CDC and SQL Replication works in such way that every data processed in an UOW on System A is transferred from A to B immediately after a commit is issued. You only need to define the tables involved in each subscription (where each subscription monitors and handles a set of tables that takes part in a UOW).
On the other hand, then Federated connection is only suitable for batch-like processing (scheduled and/or event driven) of data between A and B.
You need to make sure that the workload (by looking at how your system works) on instance B (the target) handles the chosen solution of data transfer in a correct manner.
That is. Ask yourself these questions:
Does system B need data immediately after UOW ends on A ?
Does system B need data on a scheduled/event driven basis ?
Regards,
Tomas Helgi
ASKER
Hi Tomas/ All
Thanks for inputs.
I am responding for the below Questions.
Does system B need data immediately after UOW ends on A ?
Does system B need data on a scheduled/event driven basis ?
Target System (System B) does not require the data immediately , i require daily once.
But I do not have any Tracking column ( like modified_date) to track the changes(deletes) or updates in the table.
As I do not have tracking column, thinking of opting for SQL replication instead of Federation (with PL/SQL procedures for data copy).
Thanks for inputs.
I am responding for the below Questions.
Does system B need data immediately after UOW ends on A ?
Does system B need data on a scheduled/event driven basis ?
Target System (System B) does not require the data immediately , i require daily once.
But I do not have any Tracking column ( like modified_date) to track the changes(deletes) or updates in the table.
As I do not have tracking column, thinking of opting for SQL replication instead of Federation (with PL/SQL procedures for data copy).
ASKER
Hi Thomas
I have gone through the below documentation briefly.
https://www.ibm.com/support/knowledgecenter/en/SSTRGZ_10.2.1/com.ibm.swg.im.iis.db.repl.sqlrepl.doc/topics/iiyrscncsqlreplovu.html
Can you please help me to have the direct steps with commands to simplify the things to set up the SQL replication Process.
I have gone through the below documentation briefly.
https://www.ibm.com/support/knowledgecenter/en/SSTRGZ_10.2.1/com.ibm.swg.im.iis.db.repl.sqlrepl.doc/topics/iiyrscncsqlreplovu.html
Can you please help me to have the direct steps with commands to simplify the things to set up the SQL replication Process.
ASKER
Hi Thomas / All
Good Morning!!!
I read the below passage in the IBM documentation(
https://www.ibm.com/support/knowledgecenter/SSTRGZ_10.2.1/com.ibm.cdcdoc.sysreq.doc/concepts/supportedsourceandtargets.html
)
"Each database that will act as a source for replication will need to have an instance of the InfoSphere CDC replication engine installed, as will each target for replication.
Both the Management Console and Access Server applications must also be installed in order to configure and monitor replication between the source database and the target of replication."
I am using DB2 10.5 LUW, Do i need to install InfoSphere to set up the data replication? I actually want to do with out GUI, though
commands.
Can you please let me know whether i can achieve data replication with out GUI set up?
Thanks in advance.
Good Morning!!!
I read the below passage in the IBM documentation(
https://www.ibm.com/support/knowledgecenter/SSTRGZ_10.2.1/com.ibm.cdcdoc.sysreq.doc/concepts/supportedsourceandtargets.html
)
"Each database that will act as a source for replication will need to have an instance of the InfoSphere CDC replication engine installed, as will each target for replication.
Both the Management Console and Access Server applications must also be installed in order to configure and monitor replication between the source database and the target of replication."
I am using DB2 10.5 LUW, Do i need to install InfoSphere to set up the data replication? I actually want to do with out GUI, though
commands.
Can you please let me know whether i can achieve data replication with out GUI set up?
Thanks in advance.
Hi Prardhan,
DB2 will replicate to other DB2 servers without running Infosphere.
Here's an older article on setting up DB2 replication. It's worth a read...
https://www.ibm.com/developerworks/data/library/techarticle/dm-0405nikolopoulou/index.html
DB2 will replicate to other DB2 servers without running Infosphere.
Here's an older article on setting up DB2 replication. It's worth a read...
https://www.ibm.com/developerworks/data/library/techarticle/dm-0405nikolopoulou/index.html
ASKER
Hi All
What is the difference between SQL replication and Q replication?
Which technique is best suitable for OLTP Data bases.
What is the difference between SQL replication and Q replication?
Which technique is best suitable for OLTP Data bases.
ASKER
HI All
I am facing the below error while creating the registration with below ASNCLP command:
CREATE REGISTRATION (schemaname.tablename) DIFFERENTIALREFRESH STAGE cdtablename;
ASN1550E The replication action "connection object" ended in error. The value for the input parameter "server information" is missing.
ASN1954E ASNCLP : Command failed.
I am facing the below error while creating the registration with below ASNCLP command:
CREATE REGISTRATION (schemaname.tablename) DIFFERENTIALREFRESH STAGE cdtablename;
ASN1550E The replication action "connection object" ended in error. The value for the input parameter "server information" is missing.
ASN1954E ASNCLP : Command failed.
ASKER
I have already altered the needed tables with below SQLS
ALTER TABLE schemaname.tablename DATA CAPTURE CHANGES;
And
I have also created the Control tables with below ASNCLP commands. Creation of Control tables is successful.
CREATE CONTROL TABLES FOR CAPTURE SERVER IN UW UOW TABSPCE16K OTHERS TABSPCE16K;
But I am not sure, why I am facing the below error while creating the Registration:
ASN1550E The replication action "connection object" ended in error. The value for the input parameter "server information" is missing.
ASN1954E ASNCLP : Command failed.
ALTER TABLE schemaname.tablename DATA CAPTURE CHANGES;
And
I have also created the Control tables with below ASNCLP commands. Creation of Control tables is successful.
CREATE CONTROL TABLES FOR CAPTURE SERVER IN UW UOW TABSPCE16K OTHERS TABSPCE16K;
But I am not sure, why I am facing the below error while creating the Registration:
ASN1550E The replication action "connection object" ended in error. The value for the input parameter "server information" is missing.
ASN1954E ASNCLP : Command failed.
Hi,
Make sure you define CAPTURE SERVER and/or APPLY SERVER using the SET command shown in the example in second link
https://www.ibm.com/support/knowledgecenter/en/SSTRGZ_10.2.1/com.ibm.swg.im.iis.db.repl.sqlrepl.doc/topics/iiyrsctbcrtluw.html?cp=SSEPGG_10.5.0
https://www.ibm.com/support/knowledgecenter/en/SSTRGZ_10.2.1/com.ibm.swg.im.iis.db.repl.asnclp.sql.doc/topics/iiyrsclpbldscriptsql.html?cp=SSEPGG_10.5.0
Regards,
Tomas Helgi
Make sure you define CAPTURE SERVER and/or APPLY SERVER using the SET command shown in the example in second link
https://www.ibm.com/support/knowledgecenter/en/SSTRGZ_10.2.1/com.ibm.swg.im.iis.db.repl.sqlrepl.doc/topics/iiyrsctbcrtluw.html?cp=SSEPGG_10.5.0
https://www.ibm.com/support/knowledgecenter/en/SSTRGZ_10.2.1/com.ibm.swg.im.iis.db.repl.asnclp.sql.doc/topics/iiyrsclpbldscriptsql.html?cp=SSEPGG_10.5.0
Regards,
Tomas Helgi
This question needs an answer!
Become an EE member today
7 DAY FREE TRIALMembers can start a 7-Day Free trial then enjoy unlimited access to the platform.
View membership options
or
Learn why we charge membership fees
We get it - no one likes a content blocker. Take one extra minute and find out why we block content.
Are you running DB2 systems that are licensed for Federation? That will affect your options.
If so, the easiest solution might be to query the tables as if they were in different schemas in the same instance.
How do you determine what rows will need to be copied? Do the tables have "Created Date" and/or "Modified Date" columns? How about an identity column that can be used as a marker for existing/new data? (Though that won't catch changed data.)
Will changed rows need to replace rows on the second system?