Jason's Blog on Oracle Modernization Solutions

Sorry, I've been a bit slow getting posting up recently. Finally the book is done. My last blog I announced the publication, but there was still a lot of last minute crunch time editing. Happy to say that the ink is wet, but will be dry by Open World. You will be able to grab a copy at the book store there, or from all the good online bookstores. (amazon.com uk, dk, de, fr...etc for you amex users) http://www.packtpub.com/oracle-modernization-solutions/book

Anyway...done plugging.

So, getting a lot of Sybase migration questions these days. There are a couple of ways that we can look at doing this, and it depends on what the requirements are, of course.

You can use SQL Developer and use some of the updated technology we had from the migration workbench to port that code over. That is the 'cheapest' way to do this. It can parse and and understand the Stored Procs much better than it used to.

Some of the customers have an issue of the actual time it will take to do the port..aka, the batch window from switching one db off, so the migration and sql loader path won't work because the database is just too large and there is only a few hours to port it over.

This is when we start looking at some of our partners and leverage some cool replication technologies. Yikes running out of time. Let me give you some of those details in the next installment....this week I promise

I'm happy to say that after a LOT of work, Tom Laszewksi and I have finished our book on Modernization Solutions to Oracle.

http://www.packtpub.com/oracle-modernization-solutions/book

In this book, we cover SOA Integration of the Mainframe and Re-architecture from Legacy Cobol to Open Systems using Oracle and partner tools.

What's cool about this book is that it has some hands on examples. There are no books out there like this right now, and the ones currently published are a bit dated.

To steal from the book portal:

This book combines case studies with practical examples of how to implement modernization techniques using Oracle (and partner) products to modernize to the Oracle Platform. The book also weighs the pros and cons of specific modernization use cases. Finally, we explore some of the emerging trends in technology and how they apply to legacy modernization.

So...check it out.
It should be at the book store at Open World as well.
Happy reading

Tuxedo for batch

We know that Tuxedo gives Oracle a great advantage for rehosting IMS and CICS COBOL from the Mainframe on to Open Systems. With TUX we can now provide Mainframe Quality of Service on the entire stack for CICS and IMS. So that means, our customers can gain a lot of savings from high cost hardware and OS from Z/OS by running on Linux (Even z/linux too) while opening up more standards...etc. All the good stuff we talk about.

What if you want to rehost batch programs? Is Tux good for that too? You may think the main advantage of using Tux for rehosting is that it provides the ability to do transaction processing and two phased commits, but for batch you can just use a standard COBOL compiler for linux or windows (Like MF/Accu..Fujitsu...etc)

Well, you miss out on some key RAS featuers. Let's look a bit closer.

What makes up Mainframe Batch? Let's look at this for a second. Batch on the mainframe usually looks like --> Z/OS + JES + LPRS +JCL + COBOL

Now, if you rehost batch with Micofocus Cobol only you get --> Distrubted OS (good) + JCL + COBOL

Tux looks like --> Distrubuted OS (Still good) + Tuxedo (Z/OS Like features...LPRS, JES Like runtime) + COBOL + JCL

Tuxedo provides also provides:
Connection Pooling
Clustering
Failover
Scalability
So, ask yourself, would you run a production batch system off the mainframe without these features? It's great to rehost and save money, but you CAN'T sacrifice RAS.

So here are some great reason for Batch with Tuxedo:

:: RAS (reliability, availability, scalability) built in
:: Sizable number of implementations
:: Over 2,000 implementations
:: Provides a a??Java likea?? container for COBOL applications
:: Supports C/C++ so languages like PL/1 and Assembler can be migrated to C/C++ instead of COBOL
:: SOA enabled out of the box
:: Can use any Job Scheduler you want including the free Oracle Scheduler solution

Hey Folks.
Sweden was great, btw. Customer turn out was excellent and the speakers were good. (yeah, I know, I was one of them...i meant the other guys)

So, now that Charles and Thomas have announced the Middleware story with BEA, this would be a great time to talk a bit about Oracle's Modernization offering with Tuxedo. Around here, we like to call it Re-host++. I'll post next on what I mean by that, and how we get there.
With Tux, we can look at Rehosting CICS/IMS Coobl and even manage the glut of C/C++ code out there. With clustering, failover and queing technologies, it is also and excellent solution for Batch rehosting.

Here is a link to the Tux Rehosting Page.

Tomorrow, i'll post some thoughts on why Re-hosting Batch on Tux is a good idea.
Next we'll talk about what Re-host++ is, and how we leverage a key partner, Metaware to do this.

Hey Folks:
Thought I would post about an event in Sweden coming up that we are doing with HP and Intel as a part of the Application Modernization Initiative (AMI).

The event will be 24 June, and will be in Stockholm, Sweden.

You can find details and the agenda here ->http://hp.se/evenemang/ami

There will be speakers from Oracle (me), Intel, and HP as well as industry analyst,
Phil Murphey.

I am seeing a lot of activity around Forms Modernization. I see customers wanting to upgrade forms, port them to POJO or go to something a bit more finely defined. 

The Oracle Product Management team for JDeveloper just put out this great white paper / case study on how to do forms migration.  Here Steven Price and his buddies do a couple of really cool things, besides how to get modernize forms. They highlight some key points.

1) Is goes into the notion of a phased apporach...this isn't a big bang effort.
2) It stresses the importance of legacy assets. We have business logic 'locked' up in the this legacy code, and it doesn't make sense (or efficiency) to just rip and replace it
3) It stresses the importance of the whole Oracle Stack like BPEL and Business rules.

I often see projects, be it Cobol to Java, Forms to Java or anything in between....stressing the notion of just POJO. An often you are trading one evil (monolithic, hariball code...albeit your core business), for the SAME in just a different language...sure you have classess and methods now, but it is just as ugly.

What I like about this white paper is that you can examine when it makes sense to leverage BPM tools, business rules and BPEL.  Good stuff...enjoy the read.

Legacy Modernisation  in Europe

Greetings from the UK. In the spirit of "when in rome" I'll drop my z's from this post.

I wanted to make a couple of quick observations for the level of activity I see in EMEA right now. It is very exciting, actually. I'm spending time right now the UK and France working with Oracle Partners and customers who have quite large systems that they are now looking to modernise. Notice I didn't say Migrate. A bit of a misnomer that is worth mentioning.  Folks use Modernise and Migration often interchangeably. 

Many modernisation efforts do note entail getting rid of the mainframe at all, but rather extending it via web services.  A great value-add being implemented today, which falls under this category, is extending the mainframe and integrating that with other disparate data stores and reporting off with Enterprise Business Intelligence.  With change data capture technology, we can pull data in near real time from IDMS, Adabas, SQL Server, whatever  really.  I've talked about that earlier.

Anyway...rabbit trail.
So, lots of activity here. From 10,000+ MIP Rehosts, SOA integration to automated migration of Cobol to Java....it can be found in about every country.

So, I say this to many of those who have considered these options, but don't want to be the first. That has been such a barrier for so long. I think now, we've past that moment.  In a time with there is a high demand for cost saving AND agility, the time is ripe for Modernisation.

I have a few blogs posts that i'm working regarding re-architecture that I will be posting shortly. In the mean time, I'm very behind in finishing up some chapters for a book project on Legacy Modernisation, so...until later.

I just wanted to drop a very very post here and answer the question about Z.

Can oracle run on Z?
Well, of course.
With respect to 10g, that is fully certified on Z/OS (yep...) and Z/Linux.
It is well worth mentioned as well that Fusion Middleware is also certified to run on Z/Linux as well.

I had a question from a team helping a company in Latin America doing a migration of Oracle 10g from Windows to IBM Z/Linux and was concerned about the ability of Oracle to run and support the DB and Middleware.

11g will soon be certified to run on Z/Linux.

So, short post, but hopefully will answer the question for posterity. I get the question a lot.

There are many folks who are running a LOT of COSTLY MIPS on the mainframe, yet do not want to leave the mainframe environment. (If you are reading this, then you are probably a mainframe person, so I don't need to tell you why.) There are a lot of cost savings in doing a DB2 to Oracle migration on Z/Linux. With the hot plug-able nature of Oracle Fusion Middleware, this is a very viable technical solution.

So...there ya go.

Can we run on IBM Z? Yep.

Also: You may want to check out the

zSeries Oracle Special Interest Group Conference 2008

Oracle Magazine - Legacy Modernization

Check out the new cover story in Oracle Magazine on Legacy Modernization. We were able to highlight three recent success stories and there are some nice partner highlights as well.

http://www.oracle.com/technology/oramag/oracle/08-mar/index.html

Happy New Year.
Hope all is well with you and yours. I've not written for a bit, but life has been a bit hectic recovering from OpenWorld and the end of year rush of projects.

I wanted to talk a bit about some cool things that I've been working on and seeing with respect to Legacy Modernization. As you know, there are nine ways to Sunday to "Modernization" a legacy platform. From a full replacement of the mainframe to SOA enablement.

Legacy Modernization and Business Intelligence

I'm seeing an increase of a particular implementation of SOA Services and Adapters to help extend the mainframe in the short term, while the organizations plan for larger modernizations. 

As you know, reporting and intelligence is such an important aspect of any organization. Some problems that I'm seeing a lot these days is the pain of reporting from the mainframe.
I hear managers saying that mainframe reporting is tough because:
Oracle's BI + SOA Legacy Adapters is a great answer to these problems, and is a great first step in the larger modernization story.

I've recently been on two projects, one connecting Oracle BI to AS/400 DB data to do real time change data capture from the midrange and bringing that data into a single data store for Reporting.

Another, perhaps more technically interesting, is another company that had all of the problems mentioned above, but also needed to collect data from many data stores, IMS, VSAM, Access and Oracle.

We were able to bring in one of our Modernization Alliance Partners (www.oracle.com/goto/modernization) along to do the adapters and then using Oracle ETL tools bring that into an Oracle Reporting Engine using Oracle BI.

These projects can be done in a very very short time, and provide value very quickly.  In many legacy applications, this even helped reduce MIPS (and thus save costs).

In conclusion, this can be done with pretty much any datasource you can think of, DataCom, Adabas, VSAM, DB2...whatever. Get away from MIPS intensive and old stale reports, to dynamic, near real time reporting....and you didn't migrate one line of Cobol (or Natural) code in the process

Good Stuff.