Oracle 11g Synchronous CDC & performance issues
Background.
There is approx. 2000 users and business process means there are both weekly and monthly peak loads of both data going in and and reports being pulled. The desire is for "real-time" reports of course.
I have not personally used CDC and only know what I have read via docs.oracle
Questions:
a. Is there a way to isolate CDC performance overheads? (or, failing that) Is there any generic information about the typical performance overheads of synchronous CDC available?
b.1 If the client was to move to asynchronous CDC can all the data collected go into a schema of its own (i.e. to be held completely outside the COTS application schema)
b.2 Is it possible to estimate how much delay asynchronous CDC might introduce?
b.3 Would separation of the objects for CDC into its own schema (if viable) offer possible performance advantages?
c. What alternatives to synchronous or asynchronous CDC could be proposed?
(not after full details, just hints, if necessary I'll ask more on these separately)
Large client with a Commercial Off-the-Shelf (COTS) application (J2EE 3 tiered architecture with Oracle 11g data layer)
The client has applied synchronous CDC into that COTS schema. It is not only table audits but must involve some ETL to arrive at "reporting tables".
All of this is considered a customization by the application vendor.
The client is having "performance issues" and given the presence of synchronous CDC in the schema I'm concerned about the possible performance impact CDC may be having. (Plus there a possible support issue looming over the customization.)There is approx. 2000 users and business process means there are both weekly and monthly peak loads of both data going in and and reports being pulled. The desire is for "real-time" reports of course.
I have not personally used CDC and only know what I have read via docs.oracle
Questions:
a. Is there a way to isolate CDC performance overheads? (or, failing that) Is there any generic information about the typical performance overheads of synchronous CDC available?
b.1 If the client was to move to asynchronous CDC can all the data collected go into a schema of its own (i.e. to be held completely outside the COTS application schema)
b.2 Is it possible to estimate how much delay asynchronous CDC might introduce?
b.3 Would separation of the objects for CDC into its own schema (if viable) offer possible performance advantages?
c. What alternatives to synchronous or asynchronous CDC could be proposed?
(not after full details, just hints, if necessary I'll ask more on these separately)
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER