by Sibaji Dey Choudhury

CIO expert guide to Oracle license management

Feature
Nov 11, 20127 mins
IT StrategyMobile AppsTelecommunications Industry

Oraclelicensing is complex, elaborate and has a lot of “fine print” to it. With Oracle’s never ending acquisition of technology vendors, it is often challenging for an Oracle licensing analyst or CIO to keep up with the latest licensing policies and procedures.

While many historical license metrics like Concurrent Devices and Universal Power Units (UPU) still exist in the CIO’s license inventory, many more new license metric definitions are continually being added through acquisitions.

Mostly, old licensing metrics can be translated to the latest metrics, old products can be migrated to the newer, and the more-restrictive licenses can be upgraded to the less-restrictive licences. Oracle licensing policy states that any software installed and/or used needs to be properly licensed.

Therefore there are a few questions that the CIO or IT Director must not avoid:

  1. What Oracle licenses does your company currently own?
  2. Do you have adequate Oracle licenses for all your IT environments?
  3. Do you know if your Oracle licenses are being managed effectively?
  4. Do you understand the serious implications of a contractually abiding Oracle license audit?
  5. Does your team fully understand the Oracle licensing complexities, contractual rights and restrictions?
  6. Are of aware of any exposure to financial and operational risks, by means of license over-usage or under-usage?

According to analyst house Gartner, Oracle is one of the top four software vendors conducting licence audits and at the highest frequency. Gartner also states that, if an organisation has not been audited in the last three years, then there is a 65 per cent chance that the organisation may be audited in the next 12 months.

The audit clause in the standard Oracle license agreement states that Oracle may audit the customer once-a-year upon a 45-day notice. The clause also explains that that the customer must agree to pay within 30 days of written notification any fees applicable to the use of the programs in excess of the license rights.

It is imperative for an Oracle end-user to mitigate its financial, operational and legal risks by adhering to Oracle licensing best practices. On one hand, the Oracle customer is exposed for continually paying annual technical support fees (at 22 per cent of license cost) for procured but unused software licenses, while on the other hand, the customer may still be exposed for downloading/installing excessive unpaid-for licenses and thereby causing software over usage.

As an example to prove this, let us take the example of Oracle Database Enterprise Edition (DBEE). The current license cost of one processor of DBEE is $47,500. To receive technical support and software updates, the customer also has to pay an additional 22 per cent price of the license cost, i.e. $10,450, year on year. If the customer does not use the product DBEE and yet keeps paying the technical support fees, this means that they are exposed to financial risk by losing 22 per cent of the license cost every year.

Some of the reasons why a customer becomes non-compliant on Oracle licensing are:

  • Changes to corporate structure, increased growth, mergers and acquisitions.
  • Lack of understanding of licensing policies.Hardware refresh.
  • Mixed or old metrics, as part of license inventory.
  • Installing more software than permitted by the license agreement.

It is the CIO’s responsibility to oversee that his organisation does not stand the risk of revenue leakage due to the lack of Oracle license management, during a contractually-abiding Oracle license audit. It needs to be understood that there is little or no licensing implication, across Oracle product versions like 8i, 9i, 10g or 11g.

What does Oracle vs. Rimini Street mean for the Software Support Market?

This means that when Oracle releases a new version of a product, there is no change in licensing principles. Eg., for the product Oracle Database Enterprise Edition, if Oracle releases a new database version (say 10g) from its older version (say 9i), the licensing rules do not change. Here “licensing implication” means “deviation in licensing rule”. However when the functionalities within an Oracle product family change for example in Enterprise Edition, Standard Edition, the price and licensing may be greatly affected.

Although Oracle technology is licensed at the server level and not at the instance level, yet each instance for database or middleware needs to be scrutinized because different editions and options within a product family could be installed on different instances belonging to the same Oracle server.

Some enterprise level products like Oracle Database Enterprise Edition come with options like Partitioning and Spatial, which are charged in addition to the base products. These options enhance the capabilities of the base product and need to be licensed at the same quantity and metric as the base product. Certain database options get installed automatically when downloading Oracle Database Enterprise Edition, care should be taken not to use them if not needed, otherwise they become chargeable. Features, as opposed to Options, come free of cost along with the base product. Examples of features are Dataguard and Locator.

Certain licensable smaller Oracle products may be part of a larger Oracle product, while certain licensable Oracle products are pre-requisites for another Oracle product. Oracle’s ERP (E-Business Suite) is shipped with a run-time version of database and application server, which becomes liable for full-use as soon as non-standard customisations are made.

A very important licensing consideration is for Oracle DR (Disaster Recovery) mechanisms. Typical Oracle DR scenarios are tape back-up, failover and standby. Many customers are unable to distinguish between a failover and a standby configuration; sometimes there is lack of clarity in interpreting the “10-day rule” as applied to a failover environment. As software is intangible, so the only way to quantify Oracle usage is to measure the processor count of the hardware on which the software is installed. In a multi-core environment, the total number of physical cores is summed up and then multiplied by a core-factor, which is specific to the server make and model. So, choosing the right hardware platform is critical in managing Oracle software costs.

All IT environments (production, test, development, etc.) where Oracle is installed need to be properly licensed. It is a general practice to allocate the more expensive “processor” metric in environments where the users are high or uncountable, for example on a production server or a web-facing server.

As far as server partitioning is concerned, hard partitioning is considered by Oracle for licensing purposes, while soft partitioning is not. Certain partitioning mechanisms like Sun technologies (Dynamic System Domains, Capped Containers), IBM technologies (LPAR) and HP technologies (nPAR, vPAR) are considered as hard partitioning. VMWare is considered as soft partitioning, and may imply surging bills if Oracle is installed on hardware with high processor count.

It is critical for a CIO to seriously consider Oracle license management for improved budgeting, informed decision-making and identifying potential needs in meeting regulatory standards. . There needs to be periodic exercises towards optimal license allocation & utilisation that will help in mitigation of financial, operational and legal risks, which in turn will make the organisation a much less attractive target for a formal Oracle license audit.

About the author:

Sibaji Dey Choudhuryis an independent consultant with over seven years of dedicated experience on Oracle License Management. He moved out of the LMS (License Management Services) department of Oracle Corporation as the Senior Licensing Consultant, and currently provides consulting services to partners and Oracle end-customers worldwide on Oracle license audits and license management.

Oracle fleshes out its ‘big data’ portfolio