Cut the “Muda” in Your Application Portfolio:
Achieving Leaner Data Center through Application Retirement

Application Retirement, Enterprise Applications No Comments »

While Toyota is a much maligned name today, it taught the world Lean Manufacturing drawing on the concept of eliminating waste (“muda” in Japanese).  That spawned an entire discipline of “Lean” in Operations, epitomized by Lean Six Sigma. Now Green IT pundits are espousing a Leaner Data Center through Application Retirement. There is indeed a significant muda in a Data Center: estimates have shown anywhere between 70 – 80% of available resources (time, people, equipment, and software) are in a near idle state for over 50% of the time. Much of the lean in a Data Center had, till now, focused on Virtualization and Consolidation.  Now a new branch is evolving – and it is evolving rapidly – Rationalization, a major subset of which is Application Retirement.

Come to think of it – this mirrors the stages of lean manufacturing. The first stage was Just in Time – optimizing physical inventory across the supply chain. This is akin to Virtualization that optimizes servers and storage across multiple applications they are hosting to eliminate the muda capacity at any point of time. The second stage was Reducing Set-Ups – enabling smoother set-ups and more flexible operations. This may be compared with Consolidation of servers, storage and Data Centers that reduces muda (in terms of time, people productivity and resources) when, say, a new upgrade happens in the ERP version. The US Federal Government, for example, is embarking on the most ambitious Consolidation project ever undertaken. The other – and perhaps most important element of Lean Manufacturing was Cellular Manufacturing that rationalized manufacturing processes to reduce the number of SKUs even while increasing the number of car models. Rationalization in a Data Center aims to reduce the muda of servers, storage, applications and unnecessary processes – and maybe undertaken independent of consolidation – while at the same time the CIO is delivering more value-oriented systems to the businesses.

There are many sub-disciplines within Data Center Rationalization. The one that has caught the most attention is Application Retirement. As the name suggests, it is to cut the muda out of the no-longer-required applications but retain and manage historical data and provide access to it when (and if) required for audit, compliance or litigation support reasons. It is a category within Application Portfolio Management (APM) and is going by different names: Application Decommissioning, Application Sunsetting and Application Optimization. Mirroring them in APM – to deliver the CIO’s mantra of “more” – are Application Modernization, Application Renewal and Legacy Modernization.

The history of lean manufacturing has shown that fullest extent of process improvements comes only after implementing Cellular Manufacturing and not just JIT. The history of lean Data Centers actually started with Consolidation. The movement to Global Single Instances began over a decade ago. The last five years has seen the rapid adoption of Virtualization. We expect to see Application Retirement taking centre stage over the next decade. And believe me, there’s a lot of muda out there in the application portfolio of any enterprise – sometimes running to thousands!

How Do We Make the Complex Simple?
It’s About Design Principles. And Architecture

Enterprise Applications No Comments »

We manage a complex problem. Database archiving for packaged applications can be tricky. An ERP customer has to go through many changes: customizations & localizations, patches & upgrades, occasional cross platform migrations and even significant technology changes. ERP technology moved from client-server to Internet and now it is starting to evolve to the Services Oriented Architecture (SOA).

A question often times asked is how we manage all these changes – not just for one application but for multiple applications running on multiple databases.

Take the case of patches. There would be patches for both database and the ERP. And for ERP we have minor patches as well as major patches. Minor patch would be where there are just structural changes or changes to the data type. A major patch would be where there are additions or splits to tables or formulae changes which impact data. Each of them have a certain impact to the application and could potentially have to the archived data and tables as well. How does our archiving technology handle such changes with minimal impact to the customer environment?

We realized that solving this complex problem can only be achieved through deep engineering. A solution will have to be architected grounds-up to anticipate and handle such complex changes in a manner that is simple to install, support, implement and use without causing a major disruption in the archiving process when such a change occurs.

We adopted a few design principles that are like Core Values to our Engineering.

  • Flexibility: ability to handle all data types
  • Extensibility: ability to work across multiple platforms, databases, middleware and applications
  • Scalability: ability to cover multiple data management functions
  • Openness: for easy plug-ins with partner platforms and solutions
  • Security: providing the highest levels of data integrity and security
  • Simplicity: for support, implementation and use; the complexity is only to be in our labs
  • Robustness: ability to manage change without significant re-work and disruptions

The road to achieve this was through adoption of 3-tier J2EE architecture and developing tools that confined the complexity to only our labs.

The proof of simplicity is scalability and flexibility to manage change.

Scalability: While life of Solix EDMS started with Database Archiving for Oracle e-Business Suite, we quickly expanded the footprint to cover other applications running on Oracle and non-Oracle databases. Not only that. We today cover other data management functions such as Test Data Management, Data Masking and Application Retirement and that includes platforms like AS/400 and z/OS.

Flexibility to manage change: Our customers have upgraded from Oracle e-Business Suite R10.7 to R11 and now R12 – with minor and major patches in between. And many of them were with significant new customizations. Or sometimes customizations that were dropped from one Release to the next, because the new Release had incorporated the previous customized functionality.

We achieved this in matter of just two years. Application Archiving is definitely a complex subject. But we hide that elegantly. It’s what’s hidden that intrigues all.

A.D. 2008: The Good, The Bad and The Ugly
Welcome 2009

Enterprise Applications No Comments »

It began as a good year like all New Years do. At least for Solix it did. Early on the year, we entered new markets – and continued that through the year with customer acquisitions in Eastern Europe, support of JD Edwards archiving and application sunsetting, support of DB2 on both IBM’s i-series and z/OS and strengthened our partnerships through our integrations with Oracle’s Universal Online Archive and HDS HCAP But it was a Bad Year. The US economy tumbled and the global economy crashed. Retailers have already reported the worst year in history and many lost their jobs and their homes. Bankruptcies led to the fall of many big names. For the IT industry, it was tough times as decisions took longer and there was pressure on prices and margins. It was an Ugly Year. The Mumbai tragedy was horrific. The scene of the ugliest carnage – the Taj Hotels – also happened to be one of our customers. We pray for all the victims and their families. Talking about ugliness, the Madoff scandal could not have come at a worse time.

I am sure we are all desperately looking forward to the New Year –and hoping that the worst is behind us. There’s a lot of hope and expectations. But it will be a U – rather than a V recovery, and we need to prepare ourselves for that. The new US administration, taking office on 20th January, is talking about structural changes through a “ new energy and environment plan”. The gestation period may be long but the rewards would be ever lasting. Along with many in the IT industry, we too are doing our bit in helping to create Green IT Centers and we will continue investing in this area.

We will soon announce our new Release. It will have a number of industry-unique enhancements in Data Privacy, Long Term Data Retention, support of industry vertical applications and SaaS-enabled features. We will closely follow the trajectory of Cloud Computing. Still in its infancy, we do believe the current economic climate would accelerate its adoption. Once that happens, Data Management would become a critical support function.

In a difficult year, we grew. For that we thank our customers. Many of you were great references. And that means a lot in our business. To our partners – who enabled a market reach that otherwise would just not have been possible – we thank you for your support. To the Solix team – thank you for making our customers happy.

And this is to wish all – our colleagues, customers, partners a great year ahead of us.

Does ROI Matter?
Importance of Licensing Models

Enterprise Applications No Comments »

It’s been a few years since the debate started on software licensing. And the consensus appears to be that acquiring perpetual licenses is on a downward trend. (I guess it’s still too early to call for its eventual demise). Customers are demanding fast and easily measurable ROI on their enterprise software deployments; the software industry is responding with either open source or utility-pricing.

Moving toward utility-pricing is not surprising at a time when the IT industry is going through structural changes. It started first with Application Service Providers (ASPs) or On-Demand, which some like to call; which mutated later to a Software-as-a-Service (SaaS) model. And lately there’s another trend emerging: Cloud Computing for infrastructure with a completely new player as a leader – Amazon.

All most interesting and something we had thought about when we were architecting our then new Release 4.0 back in fall 2006. We incorporated the basics to deploy our software in a SaaS environment, or at least make it possible for Managed Service Providers (MSPs) to offer data management services using our product on a rental basis on a pay-as-you-go model. The software allows metered pricing and is multi-tenanted.

Over the last eighteen months, we have developed this further and built maturity to it and now we have begun offering on a utility-pricing model, with one key difference. For a utility, you pay more as your consumption grows even if it is sometimes on a reducing rate per unit. In our case, the more you archive the less you pay – in absolute terms.

It’s a pricing model we believe would be attractive in current economic conditions and would be trend setter for rest of the computing industry, perhaps even for the utilities.

For us, it provides a predictable and steady cash flow rather than a series of bell curves of peaks during Quarter-ends and troughs in between. Having doubled our market share in the last two years, we believe utility-pricing will help us double our market share again –this time in 12 months; but more importantly would offer our customers a predictable ROI with their operating budgets when getting capital budget approvals is getting increasingly difficult.

Does ROI Matter?
Why Aren’t You Archiving?

Enterprise Applications No Comments »

Nicholas Carr presented a thesis five years ago that IT no longer provided the differentiating competitive advantage and therefore no longer mattered. Since then, he has been proven wrong many times over with billions of dollars spent on new IT initiatives, simply because there was a positive, if not high, Returns on Investment (ROI) on most enterprise-wide IT projects.

There is always a scrutiny every time a CIO presents a project for approval. Being a relatively new discipline, Database Archiving had been one area that got relegated to a lower priority simply because there were few implementations and even fewer evidences of hard numbers on ROI.

A few years ago, at an Oracle Conference, a customer hammered “Why Aren’t You Archiving”? He presented a convincing case with a strong ROI. As with all new technologies, there’s a tipping point for adoption to accelerate. That tipping point may have just happened.

Gartner’s June 2008 Report on Hype Cycle for Storage Software Technologies has unequivocally stated that the ROI for implementing a Database Archiving solution is “exceptionally high”. It further goes on to add “When database archiving is used for application retirement (with retired data archived to tape in an XML format), the business impact can be even more dramatic”.

This Report could not have come at a better time. With the economy near to a recession, it underscores what we have been saying all along. Archive to not just improve application performance; archive to improve business agility. Archive to not just reduce storage costs; archive to reduce energy costs. Archive to not just ensuring data retention; archive to insure against an expensive litigation. The high ROI is indisputable.

And for applications retirement, the Solix archiving solution with its core as a metadata based repository, we not only move legacy data to a XML format, but we can also move it to the new database of the new application, before we finally retire it to a Tier 3 storage. Archiving before upgrading or migrating reduces time, implementation costs and carrying over “dirty” data.

Mergers and Applications Retiring

Enterprise Applications No Comments »

A Depository, implementing anti money laundering laws, recently sent rejection letters to many capital market investors who have to be registered with one of the two depositories in India. The reason given: computer glitch. Isn’t that familiar? We heard the same thing when one of the largest Banks acquired a credit card issuer in 2005. A number of affected customers even blogged about their bitter experiences. The Bank’s IT problems started earlier when it had acquired another Bank in 2004, and they were struggling with integrating the systems, which first meant a rip-and-replace of the acquired Bank’s system. The subsequent acquisition of the credit card issuer compounded the IT integration problem.

On the other hand, last week, I heard a completely different story after visiting a tax collection department. A merger of two municipalities and adding more sources of tax collections meant their systems needed change. No glitches this time. There wasn’t a single complaint of citizens getting a wrong tax notice or a valid refund claim being rejected.

This got me thinking. Why should a Bank or a Depository – with vast resources and experience in managing large systems – run into a problem while a government agency with limited capital budgets and a bureaucratic set-up get it right?

I don’t believe there’s a simple answer to this. But the tax collection agency’s approach may be revealing. They adopted the principle of application modernization rather than rip-and-replace. It may surprise a few, that a government organization is an early adopter of application modernization; not so surprising if one considers that a rip-and-replace is always more difficult for them from budget approval aspect. This tax collection agency over the years has been modernizing what was once a client-server application to one that is web-driven and event-driven; their entire business process began changing from late ‘90s with the advent of the Internet, merger of two municipalities and adding more sources of tax collection. They never had an application switch over downtime; they did not have to go through a BPR or user retraining. For most, it remained business as usual.

The reason for complexity which perhaps explains the difference as well between the mergers of the two banks and the municipalities is that commercial organizations in the same industry and region are likely to have completely different business processes. On the contrary, Government departments in the same state and country are likely to have similar business processes. In my example, while the tax departments of the two municipalities were running applications from different vendors, their integration was simpler compared to what two commercial organizations may go through even if, hypothetically, they run the same ERPs and core banking systems. If they are different, and if they run on different databases, then what the Banks did – application retirement for one – may seem near inevitable.


© Solix Technologies, Inc.
Entries RSS