Presenting ECX 2.5, Our Biggest Release Yet! Posted on Oct 20th, 2016 by Sathya Sankaran

Categories: ECX

ECX biggest release

By leveraging native copy services provided by storage, they validated our vision and value proposition even more. Within a few days of that announcement, we saw heavy interest from various storage vendors to add support for their storage arrays within our platform to provide “in-place” copy data management using storage snapshots, replication and zero-footprint clones.

The leading contenders were the Pure Storage FlashArray family and IBM Spectrum Accelerate family of Storage (a.k.a. XIV). Both platforms come with industry leading data reduction, deployment simplicity, and deliver superior performance with their flash offerings. Our copy data management features deployed on top of these platforms allow customers to squeeze every penny of value out of their investments into these great products. Pure Storage preferred to announce the partnership at VMworld and IBM preferred at IBM Edge, both only a couple of weeks apart. We’ve historically only been able to add support for one storage platform every major release, every 4 months. This time, with two willing and supportive partners, our adventurous engineering team took on the challenge to deliver support for both in the same release cycle. And we did it!

Releasing support for Oracle in ECX 2.4 showed us that the value of copy data management is exponentially higher for applications. A leading bank told us that for every 20 TB of Oracle data that they have in production, they provision over 300 TB (!!) of storage to accommodate all the copies they need for performance testing, regression testing, analytics, staging, backups, etc.

In ECX 2.4, we ingeniously came up with an agentless method (using runtime code injection) to provide storage snapshot support for Oracle. The method had near-zero impact on database servers enabling creation of much more frequent copies. Use of zero-footprint storage clones ensured one can provision many copies of these databases with no additional storage utilization. And we found a way to record our copies within RMAN and weave together data masking workflows. However, Oracle was always a proof of concept, the first of many applications we intended to support. Next on the list for ECX 2.5 was expanding the same support for Microsoft SQL Server, and we did it!

All data management solutions are traditionally handled through an operations team. We recognized a shift in more organizations moving to a self-service model – letting business units take care of their own needs while still retaining central control within IT. In order to aid this, we wanted to expand our RBAC capabilities to include fine grained control over not just policies, workflows and reports, but also individual VMs and databases. And we did it!

And there is an even bigger transition of organizations moving away from a model of people running systems to a model of systems running systems. This is an essential part of continuous integration and continuous development (CICD), a DevOps model. There are popular tools such as Jenkins and Puppet that specialize in this world of spinning up test servers with static images, but struggle to mirror and keep up with production datasets. An even bigger struggle involves transitioning of these test setups into production, or dealing with data clean up. This is our bread and butter! We saw a mutual fit where we could build plug-ins for DevOps tools like Puppet and Jenkins to leverage our workflows, and we did it!

If that wasn’t enough work put into ECX 2.5, late in the release cycle a prestigious storage vendor (still can’t name names) showed interest in OEM’ing our solution. In record time, we refactored our framework to allow for white-labeling the solution. While we did deliver above and beyond all our targets in ECX 2.5, perhaps the best compliments came from a major player’s decision to put their name on our product.

It’s really quite a release!