Rubrik is not an unknown company to our readers as I already reported twice on this interesting startup. The first time I met with Rubrik it was during Virtualization Field Day in Boston I reported about their efficient data management architecture. The second time it was during VMworld Barcelona where you could read the post and watch the video interview with Chris Wahl. Today the announcement will be about the evolution and new features of their already third release called Rubrik Firefly. This release adds physical workloads, extends to edge environments and spans across clouds. Rubrik Data protection architecture is combined hardware and software package destined for VMware vSphere environments and now also for physical workloads.
Rubrik uses Atlas file system:
Almost all operations in our system involve operating on the content of a VM. In order to make these operations efficient, Atlas places entire file replicas on specific nodes, rather than spreading the data blocks backing the file randomly throughout the system. Put another way, for a file with three replicas, three nodes in our system will each contain the entirety of the data representing that file. This allows for us to run data management jobs for files on any of these three nodes with the data completely local.
Rubrik's Data protection infrastructure discovers the infrastructure automatically. When Rubrik first connects to the environment (needs vCenter connection and credentials details) all environment (with hosts, VMs and applications that run in those VMs) is discovered. Rubrik uses a policy-driven, flash ingest, and cluster-wide data reduction architecture.
The architecture is:
- Flash-Optimized – The system is built for a hybrid flash/disk architecture to maximize I/O throughput.
- Storage Efficient – The system utilizes zero-copy clones to make multiple copies of data from one “golden image.”
- Scale-out NFS server – The system exposes itself as a scale-out NFS server to any host when a snapshot is mounted.
Rubrik can restore VMs instantly, allowing to run those VMs directly on the Rubrik's storage where each “brick” can give between 10-20 thousand IOPS.
Rubrik is a relatively new company, but people especially like the:
- Easy setup – automated discovery of the whole architecture
- Simple to scale – can start small, and grow in 2U increments. The filesystem can scale and can run on white label hardware like Google or FB
- Lower TCO – data center footprint is smaller. (by 70%) and savings are between 30-50% immediately.
Customers are building cloud-based environments, but still using legacy protection architectures with traditional backups using snapshots to protect VMs. Rubrik does not. Basically all is policy driven. You assign policies and just track the compliance. You have instant search across the whole architecture and manage the data everywhere.
Rubrik is a programmatic software fabric which is based on APIs.
Rubrik can orchestrate the data retention across different clouds. Previously only AWS was supported, but in this third release, there is also Microsoft Azure. This allows Rubrik to be cheaper than tape.
It’s possible to pull a single file from Amazon AWS without the need to pull the whole VM (VMDK) as this costs you more money. Amazon S3 does charge on the volume pulled down, so if you only need to recover few files within a VM without the need to download the whole VMDK, it’s rather cost-effective.
Rubrik Firefly – 3rd Release of Rubrik:
So what's new in this release?
- SQL Server (Physical) – Forever incremental backups and automated database discovery for physical SQL servers. SLA policy engine delivers point-in-time recovery with intelligent log management. Installation of an “agent” , a connector service is necessary. (msi package). Normal daily backups, incremental backup of logs. You can keep the logs on the server until Rubrik initiates a pull request.
- Linux (Physical) – can protect Linux workloads. It uses policy-driven automation to execute granular file-set protection and recovery. It's possible to protect by snapshot let's say the /home directory. You can then restore, export or download. It has automated host discovery.
- Edge Virtual Appliance for ROBO – you can deploy edge appliance (OVA file) at ROBO office, setup data protection. Set the backup policy for your VMs at your remote office. You can also replicate those backups back to the headquarters. The size depends on the sizing, and how many VMs you want to protect.
- Azure as an archive – this is new. With a simple radio button, you can set up an archival location. Microsoft Azure blog storage is now available. For the config, you need few details like Account key, container or RSA Key, but it's similar as for Amazon AWS. You'll get global deduplication, global search, security and SLA based tiering capabilities for this storage location.
- Erasure Coding – increase storage utilization and efficiency while delivering industry-leading fault tolerance and performance. It's also performing faster because the system is designed for reads and writes.
Rubrik Hardware Appliance Specifications:
Hardware and Software, all bundled in. All data are deduplicated, compressed before written to storage. Efficiency seems to be one of the keys which by using global deduplication, allows maximizing the capacity savings across the stack.