Invalidity dossier

US 8332844

Root image caching and indexing for block-level distributed application management

Current assignee: Health Care Service Corp

Added 4/27/2026, 7:39:12 AM

Active provider: Google · gemini-2.5-flash

Auto-generating section 1 of 2: Extensions

Each section takes ~30-60s with web-search grounding. Keep this tab open — sections will fill in below as they complete.

Patent summary

Title, assignee, inventors, filing/issue dates, abstract, and a plain-language overview of the claims.

✓ Generated

US Patent 8332844, titled "Root image caching and indexing for block-level distributed application management," is currently active and is set to expire on April 6, 2028.

Here is a concise summary of the patent:

  • Patent Number: US8332844B1
  • Title: Root image caching and indexing for block-level distributed application management
  • Current Assignee: Intellectual Ventures II LLC
  • Original Assignee: Emendable Assets LLC
  • Inventors: Pradip Kulkarni, Mukul Kumar, Adhir Potdar, Richard Au, Tung Nguyen
  • Filing Date: February 21, 2007
  • Issue Date: December 11, 2012
  • Abstract: The patent describes technology for managing distributed applications at the block level, focusing on root image caching and indexing. It involves storing blocks of a root image on a primary storage unit and storing blocks of individual "leaf" images on separate secondary storage units for each compute node. These leaf images only contain data blocks not present in the root image or changes made by the respective compute nodes to the root image blocks. The technology also includes caching blocks of the root image that have been accessed by at least one compute node and receiving indexing results for the root image from one compute node to then provide those results to other compute nodes.

Plain-Language Overview of Independent Claims:

  • Claim 1 (System for Data Provision): This system provides data to multiple computers (compute nodes). It includes a main storage unit for a base software image (the "root image") divided into blocks. Each compute node also has its own storage unit for a "leaf image," which only stores new data blocks or changes made by that specific compute node to the root image, without duplicating any unchanged root image blocks. Additionally, a cache memory stores parts of the root image that have been previously accessed by any of the compute nodes.

  • Claim 7 (Method for Data Provision): This method involves several steps to provide data to multiple compute nodes. It includes storing blocks of a root image on a first storage unit and storing individual leaf images for each compute node on separate second storage units. These leaf images are designed to contain only additional data blocks not found in the root image and any modifications made by the compute nodes to the root image blocks, intentionally omitting unchanged root image blocks. The method also incorporates caching blocks of the root image that have been accessed by at least one compute node in a cache memory.

  • Claim 14 (System for File System Indexing): This system is designed to index file systems across multiple compute nodes. It features a first storage unit for a root image and multiple second storage units, each storing a leaf image for a respective compute node. The leaf images, as described previously, only contain additions or changes made to the root image, not unchanged root image blocks. A key component is the "union block devices" (UBDs) which interface between the compute nodes and the storage units, merging the root image blocks with the respective leaf image blocks to present a complete file system to each compute node. Crucially, at least one compute node is configured to index the root image, and the results of this indexing are shared with other compute nodes.

  • Claim 19 (Method for File System Indexing): This method details how file systems are indexed for a group of compute nodes. It involves storing root image blocks on a first storage unit and leaf images (containing only new or changed data, not unchanged root image blocks) on separate second storage units for each compute node. The root image and leaf images are merged to create the individual file systems for each compute node. The method then includes receiving the indexing results for the root image from one compute node and subsequently providing these results to the other compute nodes in the group.

  • Claim 23 (Computer-Readable Storage Medium for Indexing): This claim covers a computer-readable storage medium that contains instructions. When these instructions are executed by a first compute node, they cause the device to receive data blocks forming a file system. This file system is a block-level merge of a root image portion and a leaf image portion. The leaf image portion specifically contains only new data blocks not in the root image and changes made by the first compute node to the root image blocks, excluding unchanged root image blocks. The instructions further cause the first compute node to index its root image portion and then provide these indexing results to a second compute node.

Litigation Status:
US Patent 8332844 is involved in ongoing litigation. As of the current date, there are multiple cases filed in various US District Courts, including the Texas Western District Court, Delaware District Court, Texas Northern District Court, Tennessee Western District Court, Massachusetts District Court, and Texas Eastern District Court. Additionally, there have been several PTAB (Patent Trial and Appeal Board) cases filed, including IPR2025-00931 (Not Instituted - Merits), IPR2025-00840 (Not Instituted - Procedural), and IPR2025-00202 (Settlement).

Generated 5/29/2026, 2:07:37 PM