Version control a multi-terrabyte database?! Yes !


Can you version control database?

How do you manage databases in application development?  For all the advances in application development such as machine virtualization , agile development, devops practices, the database still stands as a huge barrier to development agility.

For code  management  it’s easy to use source control such as git, svn, perforce etc and for provisioning environments it can be accomplished efficiently with chef, puppet,vagrant,  jenkins. The tools support source control code, spinning up environments and testing via continuous integration and continuous delivery.

But what about the database? Databases can be in the multi-terrabytes. How does one provision environments with multi-terrabyte databases and version control these databases as changes are made by multiple different developers and as multiple versions of code and database states need to be maintained?

Answer: companies don’t version control databases. They suffer. They suffer thinking that it’s just the way life is. They suffer the consequences of lack of databases for development environments and lack of version control for the databases. The result is more bugs and most importantly slow application development times. But ..

There is a solution. The solution is data virtualization. Data virtualization empowers companies to spin up even multi-terrabyte databases in minutes for almost no storage overhead and it allows them to easily tag, branch and version control their databases.

On average with the solution of Delphix data virtualization and DB Maestro code management tools we see a 2x improvement in application development output. It’s like taking a team of 100 developers and getting the output of 200 developers. It’s actually better than that because just adding more people to a project doesn’t necessarily make the project go faster where as with Delphx and DB Maestro we do.

Find out more in a webinar Manage Database like Codebase with

  • Tim O’Brian analyst at Gleanster
  • Kyle Hailey of Delphix
  • Yaniv Yehuda  of DB Maestro

Industry analyst Tim O’Brian will share insights into problems and solutions they have analyzed in the industry and then Kyle Hailey and Yaniv Yehuda will explain how the new technologies of data virtualization and database code control work together to eliminate bottlenecks and problems in application development projects that work with databases.

Register Here


Companies rely on applications to supply the information that business uses to make decisions that generate revenue. Companies are constantly improving applications or creating new applications.
Application development depends upon provisioning environments for developers and QA.

The slower and more costly provisioning development and QA environments the more delays, bugs there will be in applications and the less revenue businesses will generate.
Businesses are constantly trimming IT budgets which leads to slower and few environment being provisioned which means more delays in the applications that the business depends on to generate revenue which in turn puts more pressure to trim IT budgets.
As examples
  • costly and slow environment provisioning means most of QA costs are spent on provisioning instead of actually QA’ing applications
  • Slow QA environment provisioning means it takes longer to find bugs and the longer it takes to find bugs the costlier it is to fix them (and sometimes the bugs actually have to go do production because of time pressures)
  • Developers sharing environments means developers block each other over those shared resources slowing down development
  • Developers who have their own synthetic , non-representational of production, environments create bugs that are not caught until late in the development cycle or even in production.
  • slow environment builds in general mean delays
In an interview with Gene Kim, author of The Phoenix Project, he said

“One of the most powerful things that organizations can do is to enable development and testing to get environment they need when they need it“

Delphix agile data platform transform environment provisioning from an expensive slow process that requires days or weeks and teams of people and TB of disk to an automated, self service process that can be done in minutes for almost no storage overhead.
With Delphix, all developers  can be given full size copies of the production database and make changes as fast as they want.

The trouble is how does one integrate all those changes?

DBMaestro tracks all the changes and and will merge the changes and highlight conflicts.



  1. Trackbacks

  2. No trackbacks yet.

  2. Vikram
    | #1

    I do agree to the benefits highlighted in the article, however if we look at the process we follow right now, all the java, sql, unix shell/dos batch scripts are in one place e.g. cvs. Once we start using the above techniques how do we overcome the problem of keeping the virtual db, dbmaestro and cvs in sync?

    Also, will there be an overhead of extra resources needed to manage the extra environment etc.

  3. LL
  4. khailey

two × = 8