Rubrik Inc v. Commvault Systems Inc

1. Case Identification

• Case #: IPR2021-00609
• Patent #: 10,210,048
• Filed: March 10, 2021
• Petitioner(s): Rubrik, Inc.
• Patent Owner(s): Commvault Systems, Inc.
• Challenged Claims: 1-20

2. Patent Overview

• Title: Selective Snapshot and Backup Copy Operations for Individual Virtual Machines in a Shared Storage
• Brief Description: The ’048 patent describes systems and methods for performing snapshot and backup operations for individual virtual machines (VMs) that are stored on a shared physical storage device. The technology purports to solve difficulties in conventional systems that perform such operations at the Logical Unit Number (LUN) level, which affects all VMs on a LUN, by enabling operations with VM-level granularity.

3. Grounds for Unpatentability

Ground 1: Claims 1-2, 4-6, 8, 10-12, 14, 16, 18-19 are obvious over Hiltgen, Vaghani, and Boda.

• Prior Art Relied Upon: Hiltgen (Patent 9,354,927), Vaghani (Patent 8,677,085), and Boda (Patent 9,639,428).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that the combination of Hiltgen, Vaghani, and Boda teaches all limitations of the challenged claims. Hiltgen, a VMware patent, discloses a virtualization system with a "virtual server agent" that commands a hypervisor to perform snapshots. Crucially, Hiltgen teaches provisioning "network storage units" that uniquely correspond to a single VM, enabling VM-specific operations. To the extent Hiltgen is not explicit enough, Vaghani (another VMware patent) was argued to disclose the same one-to-one correspondence using "virtual volumes" (vvols), where each vvol corresponds to a virtual disk of a single VM. Boda, an EMC patent, was asserted to provide details on a data protection agent that selects a specific VM from a list for backup and snapshotting, and further discloses receiving and storing metadata (like a snapshot identifier) corresponding to the completed operation in a database.
  • Motivation to Combine: A Person of Ordinary Skill in the Art (POSITA) would combine Hiltgen and Boda to implement a known backup and recovery solution for a VMware environment, as EMC's systems (disclosed in Boda) were adapted for VMware technology. A POSITA would have been further motivated to incorporate Vaghani’s teachings on vvols to predictably improve storage system performance and achieve the well-understood goal of VM-level granularity for snapshot operations, a problem both VMware and EMC were actively trying to solve.
  • Expectation of Success: Petitioner asserted a POSITA would have had a reasonable expectation of success because Hiltgen's system was designed to be compatible with backup agents like those in Boda. The combination involved integrating known, compatible technologies (VMware and EMC) using routine programming techniques to solve a recognized industry problem.

Ground 2: Claims 3, 7, 9, 13, 15, and 20 are obvious over Hiltgen, Vaghani, Boda, and Pawar.

• Prior Art Relied Upon: Hiltgen (Patent 9,354,927), Vaghani (Patent 8,677,085), Boda (Patent 9,639,428), and Pawar (Application # 2015/0212895).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground adds Pawar to the primary combination from Ground 1 to address claim limitations regarding data storage formats. Specifically, claim 3 requires a snapshot copy to be stored in its "native format," while claim 7 requires a backup copy to be stored in a "backup format which is different." Petitioner argued that while the primary combination teaches snapshotting and backups, Pawar explicitly discloses using different data formats for different types of data copies. Pawar teaches that secondary copies like "hardware snapshots" are best stored in their native format for quick point-in-time recovery, whereas backup copies are often stored in a different, non-native format (e.g., compressed or archived) for efficient, long-term storage.
  • Motivation to Combine: A POSITA would combine Pawar's teachings with the base combination to optimize the data protection system. Storing snapshots in native format would achieve the goal of rapid recovery, while storing backups in a different, more compact format would reduce long-term storage costs. This was presented as a common-sense design choice to leverage the distinct advantages of each copy type.
  • Expectation of Success: The combination would have been expected to succeed because selecting an appropriate data format for a given task (fast recovery vs. long-term storage) is a routine and trivial implementation detail for a POSITA in the field of data storage.

Ground 3: Claim 17 is obvious over Hiltgen, Vaghani, Boda, and Bayapuneni.

• Prior Art Relied Upon: Hiltgen (Patent 9,354,927), Vaghani (Patent 8,677,085), Boda (Patent 9,639,428), and Bayapuneni (Patent 9,665,386).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground adds Bayapuneni to the primary combination to address claim 17, which requires the claimed method to be performed by "a virtual machine operated by the hypervisor." Petitioner contended that while Hiltgen suggests a "superior VM" could facilitate operations for other VMs, Bayapuneni explicitly teaches this concept with its disclosure of a "service VM." Bayapuneni’s service VM runs on the hypervisor and is used to manage storage and I/O activities, including initiating snapshot operations, thereby coordinating the snapshotting functionality as required by the claim.
  • Motivation to Combine: A POSITA would be motivated to incorporate Bayapuneni's service VM concept into the Hiltgen-Boda-Vaghani system to improve flexibility and efficiency. Offloading storage management tasks from the hypervisor itself to a dedicated service VM was a known technique to optimize resource allocation and overcome inefficiencies associated with managing different backup agents across various VM types.
  • Expectation of Success: Success was reasonably expected because Hiltgen already contemplated using a "superior VM" for management tasks, and Bayapuneni provided a known, conventional implementation for such a VM. Combining the references would represent a predictable integration of a known architectural pattern to achieve well-understood performance benefits.

4. Arguments Regarding Discretionary Denial

• Petitioner argued against discretionary denial under §314(a) based on the parallel district court litigation. Petitioner asserted that denial would be inappropriate because the scheduled trial date was over 18 months from the petition filing, demonstrating that the IPR would conclude long before trial. Furthermore, Petitioner contended it had been diligent in filing the petition, and none of the prior art references relied upon in the grounds were considered during the original prosecution of the ’048 patent.

5. Relief Requested

• Petitioner requests institution of an inter partes review and cancellation of claims 1-20 of Patent 10,210,048 as unpatentable.