Michael Barz, Lead Engineer ownCloud Infinite Scale, ownCloud GmbH

sdf


Talks

State of QA at ownCloud

One of the key features of ownCloud is and has always been our strive to deliver best quality and stability. To guarantee this today and in the future we have lined up professional teams, tools and processes. This journey will never end and we need to adapt and re-evaluate our approach on a regular basis. We invite you to a backstage tour and introduce you to our team members and tools.

Recognizing a sea of spaces in storage wonderland - Part 1

OwnCloud Infinite Scale is not only the back-end for the next generation file sync & share product, but it also embraces some radical concepts on data layout and storage management. Due to the limitations of our classic back-end, we needed to explore new ways of integrating different storage technologies into a flexible but robust platform. In the past, our architecture was built on a virtual global view which was personalized for the logged-in user.
It tried to join every resource into a single view of the world. This did not reflect the view of other users and made real collaboration on shared resources cumbersome. It also created a leaky abstraction for the different underlying storage technologies, which made it very hard for contributors to master the different involved code layers. What if we could recognize a storage which provides data as a self-contained generic entity? Wouldn't that allow us to navigate safely the seas in storage wonderland?
Join us for an interactive session exploring the Spaces concept, the underlying implementation and the consequences.

Recognizing a sea of spaces in storage wonderland - Part 2

OwnCloud Infinite Scale is not only the back-end for the next generation file sync & share product, but it also embraces some radical concepts on data layout and storage management. Due to the limitations of our classic back-end, we needed to explore new ways of integrating different storage technologies into a flexible but robust platform. In the past, our architecture was built on a virtual global view which was personalized for the logged-in user.
It tried to join every resource into a single view of the world. This did not reflect the view of other users and made real collaboration on shared resources cumbersome. It also created a leaky abstraction for the different underlying storage technologies, which made it very hard for contributors to master the different involved code layers. What if we could recognize a storage which provides data as a self-contained generic entity? Wouldn't that allow us to navigate safely the seas in storage wonderland?
Join us for an interactive session exploring the Spaces concept, the underlying implementation and the consequences.