Object storage is a solution for managing large volumes of unstructured data. In the public cloud solutions such as AWS S3 and Azure Blob store provide massive scalability at predictable prices.
However, is the public cloud the right place to store all of our data? When costs, lock-in, compliance and other factors are taken into consideration, does it make sense to rent the cloud for compute and use an on-premises object store for our data? If we use this model, what are the challenges?
In this episode, Chris talks with Neil Stobart, VP Global Systems Engineering at Cloudian and Lee Biggenden, Co-founder and director at Nephos Technologies. The conversation covers real customer examples of using object storage on-premises with public cloud for compute, as well as the reverse – object storage as a golden repository and expanding to public cloud on-demand.
You can learn more about Cloudian at https://www.cloudian.com and follow Nephos on social media at @nephostech or check out their website at https://www.nephostechnologies.com.
Timeline
- 00:00 – Introductions
- 03:30 – What are the biggest challenges storing data today?
- 06:00 – Tape is not scalable – data needs to be active
- 08:15 – What real-world examples are there with object stores?
- 13:30 – At petabyte scale, on-premises object store is more practical
- 18:00 – Hybrid can mean placing data in the right place at the right time
- 24:00 – Requirements can be driven by other factors, like M&A
- 25:20 – What are the challenges in delivering a hybrid storage solution?
- 30:20 – A Data Framework becomes critically important
- 31:40 – You’re still responsible for your own data
Copyright (c) 2019 Brookend Limited, no reproduction in part or whole without permission. Episode S01E02.
Podcast: Play in new window | Download