Exploring Data Fabric and Containers in HPE DEVs new Munch & Learn monthly gatheringsJanuary 28, 2021
HPE DEV Munch & Learn technical talks are designed to provide attendees with an opportunity to learn more about today’s most popular technologies. These free, 60-minute informative meetups are sponsored by the HPE DEV Community and provide the unique opportunity to connect with some of the industry’s leading technologists.
January’s inaugural talk on data fabric with Ted Dunning and Ellen Friedman
In the January 27th Munch & Learn session, over 160 technologists attended to listen to Ellen Friedman and Ted Dunning explore the impact of a unifying data fabric and how it works. Ellen moderated the session, starting with an invitation to make these informal meetups fun, interactive, and informative. She introduced Ted Dunning, HPE Ezmeral Data Fabric CTO, who started with a business-level look at three different use cases and then quickly moved to a deeper technical level to examine how data infrastructure played a role in each one.
Ted showed a variety of processes that are handled efficiently by the data platform rather than having to be coded over and over again into applications. The examples included:
- Event stream replication from many edge sources to the core data center
- How data infrastructure handles massive scale of data mirrored from edge to core in the case of autonomous car manufacture
- How a unifying data infrastructure supports true multitenancy for a large customer
The data infrastructure used is the HPE Ezmeral Data Fabric, a software solution based on technology from the acquisition of MapR Technologies by HPE and now part of the HPE Ezmeral Software Portfolio. This ability to run very different applications – including AI and analytics, modern and legacy, containerized and non-containerized – together on the same data system is a central theme of their free, newly published O’Reilly ebook: AI and Analytics at Scale: Lessons from Real-World Production Systems.
During the talk, Ted acknowledged that people may be accepting certain trade-offs between scale, resilience, performance, flexibility, or cost that aren’t actually necessary. He pointed out that, when you release yourself from the limitations imposed by some data infrastructure technologies, you can avoid these tradeoffs, and then proceeded to dive into technical detail on how the HPE Ezmeral Data Fabric makes this possible.
While listening to the session, attendees participated in polls, had the opportunity to ask questions. In keeping with the Munch and Learn theme several of the attendees had an opportunity to share pictures of their favorite “munchies” on the accompanying Slack channel.
February talk on containers with Tom Phelan and Nigel Poulton
During the next session, February 24th, Nigel Poulton hosted Tom Phelan, HPE Fellow and CTO of the HPE Ezmeral Runtime Enterprise, who discussed container architectures and how they can leverage the Kubernetes Container orchestrator to deploy and manage stateful, as well as microservice-based, applications.
As part of the discussion, Tom described how containers require less system resources than traditional virtual machine environments, allowing applications to be deployed more easily and run on different operating systems and hardware platforms. Tom also explained how the HPE Ezmeral Machine Learning Operations package helps support collaboration between data scientists.
To view what’s up for our next Munch & Learn session, check out our schedule.