Posted in RAHİM ÇETİNEL 30-09-2019

Achieving the Unimaginable


I know that this is an unusual way to start to a post but I would like to share my experiences in one of our customers regarding how Delphix have changed the way people think.

  • Lets give a full environment to each developer and tester
  • Lets create new databases for each report that we run against the Stand-by server
  • Let developers manage their own databases

These are not the questions that are usually asked when there is a storage shortage. These are not even being dreamed when every day of operations are spend for creating, moving, updating data from one place to another with increasing ticket times and overloads.

When dealing with daily problems, putting out fires in Production and other environments, the above points don't come around. They don't even get imagined. There is just no time.

We were in the same place with this customer: moving data taking to much time, a storage of shortage making things even more interesting because everything needs to be prioritized, projects delayed, upper management constantly receiving escalations...

All this started to change after Delphix arrived.

Achievements with Delphix

After the initial deploy and configuration Delphix ingested the main systems' Oracle database which was around 5TB. It took less than 12 hours to ingest and compress this database into Delphix.

After the first database, ingest operations continued with other database systems of the customer. After 3 months there were more than 50TB of ingested database and 150TB of virtual data. These quick achievements helped us to observe the effects and gains of Delphix.

Gains with Delphix

There are different gains observed when Delphix is being used for data related operations. These include storage, time and efficiency gains.

Storage Gains

Delphix provides a gain in storage capacity since it virtualizes all the data that is being demanded and used. Going over the numbers, having ingested a 10TB database and providing 5 copies of this database means that there is 80% storage gain: 5 copies of database would require at least 50TB of storage while this can be managed by Delphix using only 10TB storage. Currently our customer is managing more than 300TB of virtual data using Delphix.

Time Gains for DBAs

Database Administrators (DBAs) aim for a stable, fast and secure database. This requires a lot of operational work. In their daily workload however, generally this may not take most of their attention. There are requests for new database instances, new database clones, new data requirements, or solving user/test caused problems in dev, test and other lower environments. These requests and tickets take most of the DBA time that might be better spent working on Production database issues, efficiencies, improvements, security and automation..

In our customer, when Delphix and virtual Databases (vDB) started being used, dev, test, and other lower environments' related requests and tickets to DBAs are dropped significantly. Production clone operations (normally taking upto 10 hours over the 5 day period) are automated to run every night. Export and import operations became a thing of past or are reserved only for the systems that are out of scope of Delphix. Overall, Delphix relieved DBAs from daily time consuming operations and improved efficiency.

Automation Gains

Delphix does not just increase the speed of moving data but also increases the level of automation capability at any customer. Providing the Hook operations during vDB operations (start, stop, before, after...) allowed us to create, modify, fine tune automation scripts that are doing range of operations: database updates, schema drops, user and password operations, data changing and masking, etc.

Masking Gains

Delphix is not just a virtualization system. It's a Dynamic Data Platform. As a platform Delphix has a Masking Engine with extensive masking capabilities and wide range of support including but not limited to Oracle, SQL, DB2, files.

KVKK (Kişisel Verilerin Korunması Kanunu) is the GDPR equivalent Turkish law that has been in force for more than 2 years now.

Keeping up with KVKK compliances can only be possible with a capable Masking Engine and Delphix provides this as part of its Dynamic Data Platform solution.

Instant Data

Providing instant data on demand to anyone is one of greatest gains for our customer. There are two aspects to this gain: First, it is possible to provide Production data to anyone who needs it in minutes. This is mostly used for BugFix, reporting and similar requirements. Second, it is possible to Provide a prepared, masked copy of Production data. The data is prepared in data factory. This feature is used mostly by dev, test and other teams.

No alt text provided for this image

Teams Managing Their Own Data

Delphix provides SelfService capability to anyone it requires. With SelfService it is possible a team or a user managing its own data sources: virtual databases, virtual files that are connected together and running end to end. I would like to share some of the SelfService capabilities here.

  • Start/Stop: SelfService users can manage their own databases.
  • Refresh: SelfService users can refresh their own databases whenever they want from a masked source with Production quality data.
  • Rewind: SelfService users can go back an hour ago, a day ago in just minutes, using their own SelfService interface. They do not need to open a ticket DBA or operation teams, they do not need to wait hours or days to start again. It just takes minutes.
  • Bookmark: SelfService users can take a bookmark of their data anytime they want. It just takes seconds and it is possible to return to any of the bookmark times. It is also possible to share bookmarks – imagine you can return to someone else's version of their completely independent database; for instance a test team can share a "broken" system with the development team.
  • Branch: SelService users can make any number of different "versions" of their own database in minutes, regardless of its size. They can flick between different versions. Imagine the possibilities for a data catalogue of different points in time and instantly swapping between different project versions of a single database running on the same connection string.