OWB 9i Features.

Hi All,
I am planing to prepare a comparison document for feature provided by Oracle Warehouse Builder & Ab initio ver 1.11.15 as ETL.
OWB config details
Oracle Warehouse Builder Client 9i Ver 9.2.0.2.8
Oracle Warehouse Builder Repository Ver 9.2.0.2.0.
Windows XP OS.
I believe following features are not avaialable in OWB 9i Release.
1> OWB doesn't support Array, Union, Vector handling.
2> OWB can read only ASCII Serial files & not Binary, Multifiles.
3> Global Library feature is not available in OWB. To make it clear in Abinitio there is concept of Project Parameter file which can hold Global Variables can be referred in Ab initio graph, also these variables value can be assigned by execution of Unix Shell script.
4> OWB doesn't support dynamic execution in a mapping. E.g. If a mapping consist of 5 flow operators at run time based on certain criteria certain flow operators can be disabled & will not get executed. This feature is available in Abinitio.
5> Change Manager in OWB provides limited feature for Config Management.
6> Optimization can be achieved by setting Set Based or Row Based option in configuration of OWB mapping.
7> No flow operator is available for Cumulative summary records.
8> Plug in component is required for Name & Address flow operator.
I am aware that Oracle is planing to launch a new release OWB 10 G & it may be possible that some of above features will be available in the new release.
Can someone please confirm my understanding is correct?
Thanks in Advance.
Regards,
Vidyanand

Hi Vidyanand,
I will try to address some of your concerns/questions, but may need some more information on some of them:
1> OWB doesn't support Array, Union, Vector handling.
JPD: Array's are supported in the next release (end of the year). I'm not sure what you mean with Union (I'm guessing this NOT as SQL union) and vector handling. Can you elaborate a bit on those terms?
2> OWB can read only ASCII Serial files & not Binary, Multifiles.
JPD: OWB supports the capabilities of SQL Loader and External Tables. SQL Loader should be able to handle multiple files...
3> Global Library feature is not available in OWB. To make it clear in Abinitio there is concept of Project Parameter file which can hold Global Variables can be referred in Ab initio graph, also these variables value can be assigned by execution of Unix Shell script.
JPD: We are adding both global and local variable support in OWB (end of the year). You will be able to store the variables on the platform, which of course is in the database (secure and easy to access).
4> OWB doesn't support dynamic execution in a mapping. E.g. If a mapping consist of 5 flow operators at run time based on certain criteria certain flow operators can be disabled & will not get executed. This feature is available in Abinitio.
JPD: While this sounds very interesting, I'm struggling a bit with when you would want to use this, and even more with how the flow would be linked together if I randomly switch of operators...
One of the things to keep in mind is that OWB generates code (I believe Ab Initio does not and interprets metadata at runtime). So this feature is harder to implement in a code generator. However I'm not convinced this is a crucial feature for ETL... Any thoughts?
5> Change Manager in OWB provides limited feature for Config Management.
JPD: Change Manager is intended for version management. We are adding full multi-configuration support in the next release where you can attach physical characteristic to objects. That combined with Change Manager will give you a much better configuration management tool.
6> Optimization can be achieved by setting Set Based or Row Based option in configuration of OWB mapping.
JPD: This is actually not correct. Set based vs row based will influence performance but is not intended for just that! It gives you different ways of interpreting a graph. In Row based you can do if-then code and row-by-row evaluations. Setbased has only SQL language as implementation.
You can influence performance with various parameters:
- Oracle DB Hints on both extract and load operators (in mapping that is)
- Set the parallel degree on the objects (invokes database parallelism)
- Influence the various code flavors with:
> Bulk Processing and Bulk Size for processing multiple rows in one go (for rowbased settings)
> Parallel row code (for rowbased settings)
The latter allows you run PL/SQL row based sections in parallel within the DB (transparently).
7> No flow operator is available for Cumulative summary records.
JPD: I'm not sure what this means, is this covered by the aggregator?
8> Plug in component is required for Name & Address flow operator.
JPD: OWB supports this natively. What you do need is to purchase the library files that the operator runs on and you can do that from First Logic, Trillium and Data Flux allowing you to use OWB with the market leading data libraries (or with localized ones if desired). We think this is a much stronger story than a closed box vendor specific solution. OWB is intended to work globally and we open up this interface as some vendors have better support in some regions as others.
But just to state this again, you do NOT need any plug-ins into the client tool and it is all there natively. You just have a choice of data libraries.
Match/Merge is of course also supported and this is completely native (and free of charge).
I sincerely think that OWB has the strongest data quality story in the ETL business!

Similar Messages

How to model in OWB if you have a subquery in your select statement? (11.2)

Which operator (is it the Table operator) to use if I have a subquery as part of my select statement? (The way I do it works, but is not elegant. What I do is create a table for the subquery and than join that with the other sources to point to my target.) But, would appreciate the best way with OWB's features.
Here is the entire SQL (including the subquery):
SELECT to_numbe (null), course_id
at.book, lp.NAME,
(select b.pricing from qp_pricing_attributes b
where b.list_id = ll.list_id and b.product = at.product) sign,
ll.operand AS price, lhb.comments
FROM
pricing_attribs at,
list_lines ll,
list_h lp,
list_b lhb
WHERE 1 = 1
and ll.list_line_id = atrib.list_line_id
AND ll.list_header_id = lp.list_header_id
AND lp.list_header_id = lhb.list_header_id
Thanks you.

Hi
Doesn't need to be a table in the target. Let's take a very simple example, let's say you wanted to do the following;
select ename,(select dname from dept d where d.deptno=e.deptno) from emp e
Then DEPT is your lookup table and EMP is your driving table.
1. Add EMP on to the canvas.
2. Add in lookup operator
3. Bind lookup to DEPT
4. Hit finish
5. Map EMP.DEPT to DEPT.INGRP1 (the lookup operator)
6. Edit lookup operator, on Lookup Conditions tab set DEPTNO for lookup column, and DEPTNO for input attribute
7. Lookup complete now
You can add in a target table operator and map from the source EMP table and the output attributes of the lookup operator. you can change step 4 to carry on through wizard and define lookup, but I have chosen to do 5, to automatically get the names and datatypes of the inputs. There are lot of other options now in 11gR2 in the lookup.
Hope this makes sense.
Cheers
David

Targeting Specific Partition in OWB.

Hi ,
I have a partition table with the range partition, partition based on the month.
Here this parition table is my target table and load strategy is update / insert
Can any body help me that how can I point my mapping, to target the specific partition (This partition depends on load day so I can not hard code it either) only instead of whole table ? This is required because of performance probles...
Have a great time....

Hi,
look at OWB PEL feature and/or ALTER TABLE ... EXCHANGE PARTITION (use it in custom procedures)
Regards,
Oleg

New Features for OWB 9.2

Mark Van De Wiel mentioned yesterday in this forum that OWB 9.2 is now available at http://otn.oracle.com/software/products/warehouse/content.html . From the version number it would appear that this is a major new release, and it's a separate install rather than a patch on OWB9.0.4.
Would Mark, Igor or Shauna be able to give us some background on the new features in this new release? From quickly looking at the release notes they would appear to be;
- Correlated Commit (committing changes across all targets uniformly)
- Ability to create public database links
- Direct PEL (removes the need for temporary tables when exchanging partitions)
- Enhanced Flat File Support
- Mapping Debugger
- Incorporation of ETL functionality previously found in standalone Pure*Integrate
- Metadata change management using the OWB GUI rather than using OMB*Plus
- Multiple Name and Address Software Providers
- Name/Address Wizard
- New Public API for OWB
- Better support for Real Application Clusters
- Advanced Repository Security and Audit options
- Support for MITI metadata bridges to third-party products
It would be good to know the thinking behind these and what areas they think will be particularly of interest to us. Also, has anything been removed or changed that we've got used to using?
Also, any day now (fingers crossed) the patch to make the OLAP functionality in OWB9.0.4 work (the 9i 9.2.0.3.x patch?) should be out - will this work with OWB9.2 as well?
Many thanks
Mark Rittman
[email protected]

Hi Mark,
While I'm not in your list of people to respond I figure I give it a go anyways :-)
The list of features you mentioned is a nice summary. One of the main release themes is data quality as you noticed. Apart from changes to the name&address functionality we also added the advanced matching and merging into this release. So you can use custom rules to match and merge data. This also allows you to do things like householding for customer data records. In essence we have now completed the integration of Pure*Integrate into Warehouse Builder.
The mapping debugger is another big thing. This allow you to walk through your mappings and see in detail what happes to the data.
The metadata change management is now in the UI. THis means there is now object level version management in Warehouse Builder via the UI. You can also compare versions and get a difference report.
As you mentioned there are a number of smaller and bigger features next to this. The idea for this release is to make OWB a top Data Quality product and allow you to do your ETL and DQ in one tool. Apart from that we have added usability features (debugger, N&A wizard) and we have added some features to solve specific problems.
Especially the corelated commit is interesting in that respect. You can now, in a map with multiple targets control the commit, and choose to either commit all targets or none, or commit them individually (current behavior).
As usual there is a lot more, as this is indeed a major release. Keep an eye on OTN for more collateral regarding the new features. We are currently updating the site to reflect the new stuff.
On the OLAP patch, both 9.0.4 and 9.2 will work equally well with the OLAP patch (when it is available). In 9.2 we added some small thing around composites. More about that when there is an OLAP patch to test this one on in a customer site.
I hope this answered the question, if not let me know.
Jean-Pierre

New Features in OWB 11g vs. OWB 9.2

Hi
Can you share what are the new features/functionalites present in OWB 11g as compared to OWB 9.2?
Thanks

... 11g R1 has the same functionality like 10gR2 >= 10.2.0.4.
In 11g R2 BETA you can use all databases (like Data Integrator) to run the mappings, not only oracle. And you can deploy the mappings as web services.
Regards,
Detlef

How to use Oracle 10G SQL features in OWB?

We are using OWB repository 10.2.0.1.0 and OWB client 10.2.0.1.31. The Oracle version is 10 G (10.2.0.3.0). OWB is installed on Linux.
Oracle 10G SQL offers
1) Inserting Into a Table with Error Logging
2) conditional_insert_clause, Specify the conditional_insert_clause to perform a conditional multitable insert.
Did anyone use above features in OWB10G mapping?
How do we use them?
Thanks in advance.
RI

Hi,
Just a thought on the multi-table insert. OWB generates multitable insert when one source is mapped to multiple targets only on the condition that after the splitter there should not be any OWB operator that can change the cardinality of the row-set coming out of the splitter (like joiners, filters, aggregators etc.)..
Regarding the link, I was wondering where I could find a decent documentation of the shadow table and the DML Error Logging through OWB .. Thanks for the link.
OWB error messages are basically Oracle Error Messages when you consider the code that gets deployed. Other than that VLD errors are sometimes mysterious and hard to debug.. Hope someone can give us more details on how to handle such VLD errors :)
HTH
-AP

OWB 10gR2: How to disable not licensed features/options

Hi,
Is it possible in OWB 10gR2 to disable the features of the "Oracle Warehouse Builder Data Quality Option" and "Oracle Warehouse Builder Enterprise ETL Option"?
If no is there any other way to prevent developers of using not licensed features/options?
Regards
Maurice

Hi Borkur,
An additional problem related to these options is that there are not explained at all in the user documentation. So as you say it's not easy at all to know for a developer which features he can use.
For example it's still not clear for me, didn't get any feedback from Oracle Sales, if it is allowed to define a dimension without any licensing.
Also I think that's a shame that Oracle Sales defines a new licensing model (for product which was already delayed by over 18 months) once the product has been developed. If Sales would have defined the new licensing a the begging of the development it would have been possible to implement it so that you can disable these features and also they would have documented it in the user/installation manuals.
Maurice

OWB 10gR2 Merge Feature

Hi
I am trying to create a mapping which uses the merge statement in the end.
Though I just want to have a normal insert and no update.
This is possible in 10g DB wherein we can only have an update or Insert .
However, when I am trying to create the mapping using OWB 10gR2 I am forced to choose an update and insert clause... is there any way I can get around OWB 10g to just create the merge with either Only Update or Only Insert.
Thanks
Sushant

Hi
Thanks for your replies. I do know that I could use either Insert or Update in the drop down menu. However before 10gR2 We didnt have the flexibility of using either one Insert or Update.
For More Information refer to .....
http://www.oracle-developer.net/display.php?id=310
In Oracle 10g, either the INSERT or UPDATE operations are optional (you must have one, but not necessarily both, unlike in 9i). This is a critical enhancement and one which makes the MERGE statement much more flexible. In particular, this new feature could make updates with correlated sub-queries a thing of the past as MERGE is easier to code, easier to understand, more flexible, less error-prone and faster. In the following example, we'll MERGE source into target but ignore new rows by leaving out the INSERT operation. Note how much easier this is to code and read than a "double-correlated UPDATE".
Hence I wanted to try if this is possible using the Warehouse Builder 10g.
Anyways Thanks for your Time
Regards,
Sushant

New licensing for OWB 10g R2 (Paris)

Hi,
Does anyone know how much the new licenses (per DB server CPU) for OWB 10g R2 (Paris) cost?
Is it correct that the features for the modeling of SCD 1 and 2 are not included in the basic "Core ETL Features"?
If I'm using the the SCD 2 features to develop the OWB mappings do I also need to license the "Enterprise ETL" for the production server?
Related to the previous question: for the production DB where I only need the runtime part of the repository is it required to license any options?
Regards
Maurice
PS: 2 links related to these questions:
http://www.rittman.net/archives/2006/05/owb_10g_release_2_now_availabl.html
http://www.oracle.com/technology/products/warehouse/htdocs/owb_10gr2_faq.htm#HowisOWBPackaged

Maurice, not sure what to say. If you've been using OWB, then you don't lose any functionality going to the new version, and it will be "free". You simply won't get any of the new functionality (all of the old functionality is included in the "core" features)
However, as I said in the other post, I hope Oracle reconsiders the SCD 1 / 2 / 3 licensing. To me, SCD type 2 are not "enterprise level" functionality - that is base level functionality for ANY data mart or data warehouse. I have no problems with paying for options the other ETL vendors are charging for (data quality comes to mind...), but if we deploy a large DW on a 32 processor box, paying for the Oracle licenses AND an addition $300,000+ for OWB functionality just to simplify SCD type 2, seems WAY overkill. Actually, in our project, we had pretty much settled on Oracle as the RDBMS for our DW, but if OWB is up in the air, we will probably open this up for an RFQ.
Scott

Experts VS Process Flows on OWB 10g Release 2 (Paris)

I was looking at the the new features list for OWB 10g Release 2 (Paris), and I notices the new feature called Experts. Does anyone have any insight into when an Expert would be used instead of a Process Flow? It doesn't look like Process Flows are going away, so I was trying to get a grasp of the use for Experts.
From Here:
http://www.oracle.com/technology/products/warehouse/htdocs/owb10gr2%20new%20features.htm
Experts. Experts are solutions that enable advanced users to design solutions that simplify routine or complex tasks that end users perform in Warehouse Builder. You can design these solutions in the Expert Editor which resembles the Process Flow Editor and shares many of the same commands and navigational tools.

Hi,
Just to start with the Experts capabilities, they can be used for designing schema objects.......I tried with an expert for sampling the source files, creating External Tables on top of it....validating and deploying..........itz very interesting to work with.........I am unable to see the practical use of this for end users.....
where as we all know, the PF can only be for executing the deployed objects.....
Mahesh

How to manage different load frequencies in OWB?

How to manage different load frequencies in OWB?
We are using OWB repository 10.2.0.1.0 and OWB client 10.2.0.1.31. The Oracle version is 10 G (10.2.0.3.0). OWB is installed on Linux.
We have situation to bring data feed into database using OWB on different load frequency. Eg. Daily, weekly and monthly.
Are there any builtin features/best approaches in OWB to implement the above?
Does runtime audit tables maintain such load frequency data?
If not, do we need to maintain our own tables (custom) to maintain such data?
How do we able to validate before starting ETL process with respect to previous load frequency for current load?
How did our forum members handled above situations?
Any idea ?
Thanks in advance.
RI

If you look at the OWB public function WB_RT_GET_LAST_EXECUTION_TIME you can examine the query OWB uses to return this information. It is based entirely on public views that you should already have synonyms for in your OWB user.
Now, how you want to implement logic to determine if a mapping is run is entirely up to you If you are using process flows you could embed your call to start a mapping in a custom transformation that queries the repository and starts a given mapping if it deems the correct interval to have passed. We do something similar as a means to perform a process flow restart (it will skip over any successfully completed mappings from the last run if the last COMPLETE run did not finish succesfully), so that a restart picks up where the last run failed.
The main caveat I would make to mining the runtime repository is that someone purging old data might very well purge the data you are depending on, so you might want to build a couple of tables to hold your own metadata for you. E.g. one table holding the mapping name and desired interval between loads, and another holding the mapping name and last run date. The second table could get updated by the post-mapping procedure of each mapping.
Anyway, the data IS there in the repository. How you intend to access it or how much you are willing to trust its persistence is up to you.
Mike

OWB11gR2: Mapping execution in a process flow not visible in OWB Browser

When a mapping is executed inside a process flow, execution details are not visible in OWB Repository Browser (Control Center reports) - rows processed, errors etc. Mapping row is missing in a log, like it never happened (but it did).
This auditing information is very important for monitoring reasons (to our customers also) and I just don't get it how this functionality is lost with this version. Another serious bug?

Hi David,
I was rather tired and frustrated last evening, so today I noticed some things I didn't yesterday. Your reply gave me a new motivation.
The conclusion is - a mapping execution in a process flow is logged, but the way activities are displayed in OWB Browser are now different than in previous versions. If I click on 'Execution Job Report' on a process flow, I see all the activities listed except mappings (transformations, assign, file exists, subprocess etc.). If I want to see mapping execution row, I must click on a plus (expand) sign.
This kind of behavior will make processes with a complex hierarchy (usually we have more than 5 levels of subprocesses) rather vast to monitor. In 10gR2, a drilling down was accomplished by opening a new browser tab (Execution Job Report link) for each subprocess/mapping activity. Now it shall remain on one huge screen (list) that keeps expanding.
But, if that is the new feature, we shall live with that. If our customers won't like it, they will have to get used to it.
Thank you for your reply!

OWB process flow & Oracle Workflow

Hi,
In order to deploy the OWB process flow packages, is it necessary to have Oracle workflow installed on the Target server? When I tried to deploy a process flow package to the target database (10gR2 db), it required a user called owf_mgr. As far I know, owf_mgr is the admin user for workflow. Can this feature work without OWF being installed?
Thanks for your help
regards

You also might be interested in this thread...
OWB and Work Flow (being dessuported)
Just in case you are starting an implementation.

What are the best approaches for mapping re-start in OWB?

What are the best approaches for mapping re-start in OWB?
We are using OWB repository 10.2.0.1.0 and OWB client 10.2.0.1.31. The Oracle version is 10 G (10.2.0.3.0). OWB is installed on Linux.
We have number of mappings. We built process flows for mappings as well.
I like to know, what are the best approches to incorportate re-start options in our process. ie a failure of mapping in process flow.
How do we re-cycle failed rows?
Are there any builtin features/best approaches in OWB to implement the above?
Does runtime audit tables help us to build re-start process?
If not, do we need to maintain our own tables (custom) to maintain such data?
How did our forum members handled above situations?
Any idea ?
Thanks in advance.
RI

Hi RI,
How many mappings (range) do you have in a process flows?Several hundreds (100-300 mappings).
If we have three mappings (eg m1, m2, m3) in process flow. What will happen if m2 fails?Suppose mappings connected sequentially (m1 -> m2 -> m3). When m2 fails then processflow is suspended (transition to m3 will not be performed). You should obviate cause of error (modify mapping and redeploy, correct data, etc) and then repeat m2 mapping execution from Workflow monitor - open diagram with processflow, select mapping m2 and click button Expedite, choose option Repeat.
In re-start, will it run m1 again and m2 son on, or will it re-start at row1 of m2?You can specify restart point. "at row1 of m2" - I don't understand what you mean (all mappings run in Set based mode, so in case of error all table updates will rollback,
but there are several exception - for example multiple target tables in mapping without corelated commit, or error in post-mapping - you must carefully analyze results of error).
What will happen if m3 fails?Process is suspended and you can restart execution from m3.
By having without failover and with max.number of errors=0, you achieve re-cycle failed rows to zero (0).This settings guarantee existence only two return result of mapping - SUCCSES or ERROR.
What is the impact, if we have large volume of data?In my opinion for large volume Set based mode is the prefered processing mode of data processing.
With this mode you have full range enterprise features of Oracle database - parallel query, parallel DML, nologging, etc.
Oleg

How to define a parallel, deterministic pl/sql function in OWB 11.2?

Hi,
at my customer site, we are starting with the new development work after upgrading from OWB 10.2 to 11.2 and we can't find the configuration parameter to define a parallel and deterministic function within a new package. The "old" function dialog in 10.2 had some checkboxes for those two features ...
Thanks in advance
Michael

Thanks for the advice, the Cut & Copy idea from MccM is working fine, however is not really a solution for the problem. But we are using it as a workaround for the moment.
Just for documentation: There already exists a bugfix, addressing the issue of lost parrallel and deterministic information during an upgrade from OWB10.2 to 11.2 (see bug 14673399 and patch 16082939)
And for at least some of us with access to Oracle Support: Our onsite Oracle consultant opened a service request for this issue: 3-6773545351 (You will not be able to see the request unless You are an Oracle employee, but maybe the reference is helpful for You.)

OWB 9i Features.

Similar Messages

Maybe you are looking for