HaShiCorp Inc v. Invincible IP LLC

1. Case Identification

• Case #: IPR2022-00567
• Patent #: 9,635,134
• Filed: February 14, 2022
• Petitioner(s): HashiCorp, Inc.
• Patent Owner(s): Invincible IP LLC
• Challenged Claims: 1-18

2. Patent Overview

• Title: Resource Management in a Cloud Computing Environment
• Brief Description: The ’134 patent describes methods for managing virtual machine (VM) resources in a cloud computing environment. The system monitors resource consumption rates, uses a first scheme to prioritize VMs, and then uses a second scheme to re-prioritize and migrate a VM to alternate resources outside the environment if consumption rates change and exceed a threshold.

3. Grounds for Unpatentability

Ground 1: Obviousness over Xu and Ferris - Claims 1, 5, 7, 11, 13, and 17 are obvious over Xu in view of Ferris.

• Prior Art Relied Upon: Xu ("A Multi-objective Approach to Virtual Machine Management in Datacenters," a 2011 research paper) and Ferris (Application # 2009/0300635).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Xu taught nearly all limitations of the independent claims. Xu disclosed a system for managing VMs in a datacenter that determined resource consumption rates (CPU usage, I/O rates) and used multiple resource management schemes to prioritize VMs for migration. For example, Xu’s “thermal emergency” and “resource contention” schemes prioritized VMs based on consumption rates when those rates exceeded predetermined thresholds. The only key element Xu did not explicitly teach was migrating a VM to resources located outside the current cloud computing environment. Ferris was alleged to supply this missing element, as it described a "cloud marketplace system" that explicitly migrated VMs "to a new cloud computing environment" when resources in the current environment reached a threshold value.
  • Motivation to Combine: Petitioner contended a POSITA would combine Xu’s sophisticated internal VM management with Ferris’s external migration capability. Xu’s system focused on optimizing resources within a single datacenter, but a POSITA would recognize the benefit of adding a safety valve for situations where the entire datacenter becomes overloaded. Ferris provided a known solution for this exact problem. Combining them would have been a predictable improvement to create a more robust and scalable resource management system.
  • Expectation of Success: The petition asserted that combining the teachings would be a routine implementation of well-understood resource-management functionalities. A POSITA would have a high expectation of success in applying Ferris’s concept of inter-cloud migration to Xu’s intra-cloud management system, as the underlying principles of monitoring thresholds and triggering migrations are analogous.

Ground 2: Obviousness over Xu, Ferris, and Van Riel - Claims 2-4, 6, 8-10, 12, 14-16, and 18 are obvious over Xu and Ferris in view of Van Riel.

• Prior Art Relied Upon: Xu, Ferris, and Van Riel (Application # 2007/0124540).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground built upon Ground 1 to address dependent claims reciting specific prioritization schemes, namely the "low inter-reference recency set (LIRS)" and "least recently used (LRU)" replacement schemes. Petitioner argued that Van Riel taught these well-known caching algorithms for managing virtual memory and explicitly suggested their use for allocating resources among multiple VMs. Van Riel explained that algorithms like LIRS and LRU could be used to manage memory based on recency and frequency of use, which are key factors in minimizing the cost and disruption of moving resource-consumers like VMs.
  • Motivation to Combine: Petitioner asserted that a POSITA would be motivated to incorporate the LIRS and LRU schemes from Van Riel into Xu's prioritization logic to improve migration decisions. Xu’s system already sought to balance competing objectives, such as resolving resource contention while minimizing migration costs. A POSITA would recognize that using LIRS or LRU metrics, as taught by Van Riel, would provide a more accurate estimate of migration costs and potential performance disruption. This would allow for a more refined prioritization that better achieves Xu's stated goals of efficiency and minimal impact.
  • Expectation of Success: The petition argued that applying a well-known caching algorithm (LIRS or LRU) to VM resource management was a known and predictable technique, as suggested by Van Riel itself. A POSITA would have a reasonable expectation of success in modifying Xu’s simple prioritization formulas to incorporate these well-understood metrics, as it represented a straightforward enhancement rather than a complex or unpredictable integration.

4. Key Claim Construction Positions

• "determining a consumption rate of cloud resources by one or more virtual machines (VMs)": Petitioner argued that a POSITA would understand this phrase to mean the rate at which cloud resources are consumed by the VMs, not a determination of that rate that is performed by the VMs. This construction was asserted to be critical, as the prior art (e.g., Xu) disclosed a central resource manager that performed the monitoring and determination, which aligns with the proposed construction and allows the art to be mapped to the claims.

5. Arguments Regarding Discretionary Denial

• Petitioner argued that the Board should not exercise discretionary denial under 35 U.S.C. §325(d). The petition contended that the primary references, Xu and Ferris, were not before the Examiner during prosecution and were not cumulative to the art that was considered. It was argued that these references disclosed key claim elements—such as using multiple prioritization schemes and migrating outside the cloud environment—that were central to the reasons for allowance, and therefore the Office lacked the benefit of this key prior art.

6. Relief Requested

• Petitioner requests institution of an inter partes review and cancellation of claims 1-18 of the ’134 patent as unpatentable.