The MinIO team is immensely proud of what we have been able to build in the high performance object storage, so we relished the opportunity to showcase our ever-growing functionality at Cloud Field Day in June. After beginning with a high-level overview, we got the chance to dive deeper into MinIO’s various features that put it at the cutting edge of the constantly evolving storage landscape.
After discussing Kubernetes, Multi-Tenancy, and GUIs in our first deep dive session at the event (well worth a watch–you can find that video here), MinIO founder AB Periasamy, CMO Jonathan Symonds, and engineer Daniel Valdivia explored three additional critical capabilities of MinIO that allow for functionality at any scale:
- Tiering: MinIO’s tiering of data allows the user to improve cost and performance by optimizing the data’s lifecycle management, and it can do this across storage mediums, across cloud types, and within the public cloud itself. For example, data in the “hot” tier can be stored on NVMe or SSD to maximize performance, then moved to HDDs for larger scale workloads or long-term storage as it gets colder. Alternatively, the user could tier high performance workloads in private cloud storage while moving colder data to less expensive public cloud infrastructure (while using MinIO in that public cloud to maintain access to the data), or use MinIO to determine whether data should be moved to block or object storage within a public cloud.
- Active Active Replication: Many enterprises want replication of their data across data centers, across regions, across private and public clouds, or across multiple private clouds or multiple public clouds. In traditional block and file storage systems, this is not scalable because it requires pushing the data, and therefore lacks consistency—with MinIO’s object storage built on versioning, however, this can be effortlessly done even at the 100PB+ level. This feature is becoming increasingly non-negotiable for enterprises with massive data infrastructures, and MinIO is the only vendor offering it today.
- Continuous Data Protection: CDP was never possible with block and file systems because they required snapshots of each individual transaction, which is simply impossible at scale. With MinIO, a user can go to where an object was at any point of time; since every change is automatically versioned, the concept of “recovering” a snapshot is irrelevant because it already exists.
The above features are not outright new—they have been present in block and file storage applications for a long time. Where MinIO is changing the game is in how we are seamlessly applying them to object storage for the enterprise. We have stated time and time again that enterprises will have no choice but to adopt object storage for its myriad benefits, and MinIO is the leader in scalability, performance, and flexibility on this front.
After a detailed discussion of the above from AB, Daniel demoed how the concepts actually play out in both on-premise and public cloud environments. A full explanation of these capabilities would be too complex for a quick post, and to truly understand their implications and impact requires seeing how they play out in action—therefore, we highly recommend you check out the full video below.