The Hunger Site

Friday, May 30, 2014

The case for Snap Clone over point tools

Today, I stumbled over a competitor blog, conspicuous by its factual incorrectness on Enterprise Manager Snap Clone. However, I must compliment the author of the blog, because inadvertently, he has raised a point that we have been highlighting all along. The author, with reference to Dataguard and storage technologies, argues against the cobbling of technologies together and adding another technology stack to the mix without any automated management.  
Precisely the point! In the wide realm of technologies, there are necessities and there are accessories aka nice-to-haves. The necessities are technologies that are needed anyway, such as a high fidelity, high performance storage from a reputed vendor or a good DR solution for a mission critical database environment. Similarly, for any Oracle DBA worth his/her salt, Enterprise Manager 12c is a necessity, a part of the daily life. The Enterprise Manager agent, keeping vigil on every host, is therefore not an overhead, but the representative (the "agent" in true sense) of the DBA. Deep diagnostics, performance management, large scale configuration management, patching and compliance management make Enterprise Manager the darling of any Oracle DBA. All surveys suggest that any DBA spends considerable amount of time in Enterprise Manager for performing things beyond just data cloning, so why invest in an accessory for the cloning of Oracle test databases and unnecessarily proliferate the number of point tools (and possibly several instances of them) that you need to manage and maintain, not to ignore the past history that cites that very few such point tools solved customers' CAPEX and OPEX problems over the long run. It is like using spreadsheet for expenses and ERP for all other financial tasks.This is not to suggest that these point tools do not have good, innovative features. Over my tenure in the industry, I have come across several such tools with nice features, but often the hidden costs outweigh the benefits. Our position in this aspect has been consistent, whether it is on a competitor’s tool or our own. Few years back, we integrated My Oracle Support into Enterprise Manager with the same consistent goal that Enterprise Manager will serve as the single pane of glass for the Oracle ecosystem. Same has been our position on any product that we acquire.
Snap Clone's support for Dataguard and native storage stems from popular customer demand to leverage technologies they already invested in, and not create standalone islands of automation. Moreover, several customers have voiced in favor of the performance and scalability advantages that they would get by leveraging the native storage APIs. How else would you support one of the world's largest banks, a Snap Clone customer, who performs 60,000 (sixty thousand) data refreshes per year! In any case, that should not imply that we bind ourselves to any of those technologies. We do support cloning on various storage systems based on ZFS filesystem. Similarly, the Test Master refresh can be achieved with one among RMAN, Dataguard, Golden Gate or storage replication and optionally orchestrated with EM Job System.
Enterprise Manager 12c has taken a great step in delivering features via plugins that can be revisioned independent of the framework. An unwanted side effect is that the awareness often lags what is actually supported in the latest version of the product. For example, the filesystem support was introduced last Fall. And of course Enterprise Manager 12c Snap Clone supports RAC. My esteemed colleague and DBA par excellence, in her blog has highlighted some of these to dispel some of the prevalent awareness issues. Snap Clone's usage among the E-Business Suite and Developer community does not need any special accreditation. It is heavily used by the world's largest E-Business Suite Developer community-the Oracle E-Business Suite Engineering organization itself! It is true that Snap Clone does not support  restoration to any arbitrary point in time, but then our customers and prospects have not voiced a need for it. In reality, most customers want to perform intermediate data transformation such as masking and subsetting as they clone from production to test, and Enterprise Manager 12c already boasts of sophisticated data masking technologies, again via the same interface. It also includes testing features like Real Application Testing (RAT) that can complement and follow the test database creation. Future releases of Enterprise Manager will support a tighter integration among these features.
Snap Clone is delivered as a part of the Database as a Service feature set that has been pioneering, industry-leading and getting adopted at a great pace. Little wonder that we have already received a copious amount of Openworld paper submissions on the topic. In this emerging trend of DBaaS adoption, we find no reason to fragment the tasks such as fresh database creation, pluggable database provisioning and cloning across silo'ed point tools (not to mention broader PaaS capabilities which may be needed for complete application testing). Each use case could be different but needs a single service delivery platform. EM12c is that platform for Oracle. Period. So, think twice before 'adding another technology to the mix'. You do not need to.

No comments:

Disclaimer

Opinions expressed in this blog are entirely the opinions of the writers of this blog, and do not reflect the position of Oracle corporation. No responsiblity will be taken for any resulting effects if any of the instructions or notes in the blog are followed. It is at the reader's own risk and liability.

Blog Archive