DCT

1:25-cv-01571

Primos Storage Technology LLC v. Amazon.com Inc

Log In
Key Events
Complaint

I. Executive Summary and Procedural Information

  • Parties & Counsel:
  • Case Identification: 1:25-cv-01571, D. Del., 12/30/2025
  • Venue Allegations: Venue is alleged to be proper in the District of Delaware because both defendants are incorporated in Delaware and have committed acts of infringement in the District.
  • Core Dispute: Plaintiff alleges that Defendant’s cloud storage services, including Amazon S3 and associated products, infringe five patents related to transaction-based data storage systems and data coding methods for ensuring durability.
  • Technical Context: The technology at issue concerns foundational methods for organizing, storing, and protecting data in large-scale distributed systems, which are critical to the operation of modern cloud computing platforms.
  • Key Procedural History: The complaint alleges that Amazon has been aware of the asserted patent families for years, citing the prosecution history of Amazon’s U.S. Patent No. 9,772,787, during which the examiner cited a publication corresponding to one of the patents-in-suit as prior art. Plaintiff alleges Amazon studied the specification in detail to overcome the rejection, which may be used to support allegations of pre-suit knowledge and willfulness.

Case Timeline

Date Event
2004-05-13 Priority Date for ’663, ’528, and ’344 Patents
2007-04-19 Priority Date for ’944 and ’356 Patents
2008-06-10 ’663 Patent Issued
2011-05-03 ’528 Patent Issued
2011-12-13 ’944 Patent Issued
2012-11-13 ’356 Patent Issued
2014-03-31 Amazon files application for its ’787 Patent, allegedly leading to awareness of Plaintiff's patents
2020-03-24 ’344 Patent Issued
2025-12-30 Complaint Filed

II. Technology and Patent(s)-in-Suit Analysis

U.S. Patent No. 7,386,663 - *"Transaction-Based Storage System and Method that Uses Variable Sized Objects to Store Data"*

  • Patent Identification: U.S. Patent No. 7,386,663, "Transaction-Based Storage System and Method that Uses Variable Sized Objects to Store Data," issued on June 10, 2008.

The Invention Explained

  • Problem Addressed: The patent’s background describes the limitations of then-existing file systems, which were typically "block-oriented," meaning they stored data in fixed-size blocks and overwrote data in place when modified (Compl. ¶20). This approach was described as inherently slow, prone to data fragmentation, and inefficient for the massive, distributed environments of emerging cloud storage (Compl. ¶¶18, 20-21; ’663 Patent, col. 4:4-23).
  • The Patented Solution: The invention proposes a new type of file system that combines "transaction logging" with techniques from RAID storage, such as "striping" (Compl. ¶25). Instead of overwriting blocks, new data and changes are written sequentially into a log, which is striped across an array of disks organized into "ranks" (’663 Patent, col. 4:51-59, col. 5:1-13). This system for storing variable-sized objects is managed by several key components: a "rank manager" to configure the disk arrays, a "stripe space table" to track data location, and a "mapping component" to link object identifiers to their physical addresses (Compl. ¶25; ’663 Patent, col. 5:1-36).
  • Technical Importance: This architecture was designed to provide a scalable, reliable, and high-performance file system better suited for the demands of large-scale, cloud-based data storage than the block-oriented systems it sought to improve upon (Compl. ¶27).

Key Claims at a Glance

  • The complaint asserts at least independent Claim 1 (Compl. ¶62).
  • The essential elements of Claim 1 are:
    • A transaction-logging data storage system with a data storage subsystem for storing variable-size data objects.
    • The subsystem uses storage devices comprising an array of data storage units, configured into one or more "ranks" that provide "stripes" for storage.
    • A "rank manager" performs configuration processing for the ranks.
    • The subsystem is configured to write data to the stripes as a unit.
    • A "stripe space table" tracks space usage.
    • A "mapping component" processes mapping information between object identifiers and their physical addresses.
    • A processing component coordinates these parts to implement logging of data objects and changes to them.
    • The mapping component includes subcomponents to enable the emulation of common data storage abstractions.
  • The complaint does not explicitly reserve the right to assert dependent claims for this patent.

U.S. Patent No. 7,937,528 - *"Transaction-Based Storage System and Method that Uses Variable Sized Objects to Store Data"*

  • Patent Identification: U.S. Patent No. 7,937,528, "Transaction-Based Storage System and Method that Uses Variable Sized Objects to Store Data," issued on May 3, 2011.

The Invention Explained

  • Problem Addressed: As a continuation of the application that led to the ’663 Patent, this patent addresses the same technical problems related to the performance, reliability, and scalability limitations of conventional block-oriented and journaling file systems (Compl. ¶¶15, 17-18).
  • The Patented Solution: This patent claims a method for logging transactions in the data storage system architecture described in the shared specification, rather than the system itself (Compl. ¶104). The claimed method involves storing variable-size data objects in a subsystem with ranks and stripes, tracking space usage with a "stripe space table," processing mappings with a "mapping component," and coordinating these operations to log data objects and any changes to them (’663 Patent, col. 4:51-67, col. 5:1-36).
  • Technical Importance: By claiming the method of use, the patent provides a different scope of legal protection for the same core technological innovation, aimed at enabling high-performance, large-scale data storage (Compl. ¶27).

Key Claims at a Glance

  • The complaint asserts at least independent Claim 1 (Compl. ¶104).
  • The essential elements of Claim 1 are:
    • A method of logging transactions in a data storage system.
    • Storing variable-size data objects in a data storage subsystem that uses an array of data storage units configured into ranks and stripes, with a rank manager for configuration.
    • Writing data corresponding to stripes into the ranks as a unit.
    • Keeping track of space usage via a stripe space table.
    • Processing mapping information between data object identifiers and their physical addresses via a mapping component.
    • Performing coordinated operations among these components to implement the logging of data objects and changes to them.
  • The complaint does not explicitly reserve the right to assert dependent claims for this patent.

Multi-Patent Capsule: U.S. Patent No. 10,599,344 - *"Transaction-Based Storage System and Method that Uses Variable Sized Objects to Store Data"*

  • Patent Identification: U.S. Patent No. 10,599,344, "Transaction-Based Storage System and Method that Uses Variable Sized Objects to Store Data," issued on March 24, 2020.
  • Technology Synopsis: This patent, part of the "Transaction-Based Storage Patents" family, is directed to a storage system that overcomes the limitations of prior art file systems (Compl. ¶27). The claimed solution is a system comprising a plurality of ranks with stripes for storing variable size objects, and notably includes "at least one stripe buffer for storing objects to be written into a stripe," to which objects are appended according to a "write ordering policy" (Compl. ¶142; ’344 Patent, col. 27:30-34).
  • Asserted Claims: Independent Claims 1 and 2 (Compl. ¶142).
  • Accused Features: Amazon's storage products, including S3, are alleged to infringe by using a file system that stores variable-sized objects across a plurality of ranks and utilizes at least one stripe buffer to accumulate objects before they are written into a stripe across multiple servers (Compl. ¶¶143, 146, 148).

Multi-Patent Capsule: U.S. Patent No. 8,078,944 - *"Systems, Methods and Computer Program Products Including Features for Coding and/or Recovering Data"*

  • Patent Identification: U.S. Patent No. 8,078,944, "Systems, Methods and Computer Program Products Including Features for Coding and/or Recovering Data," issued on December 13, 2011.
  • Technology Synopsis: This patent, one of the "Data Coding Patents," addresses the problem of achieving high data durability in large-scale storage systems without the high cost of simple data replication (Compl. ¶¶29-33). The patented solution is a method of transforming "W" data inputs into "H" output data streams using Galois field operations, where the transformation includes producing an "H-sized intermediary" for each input and combining them into an "H-sized result," allowing the original data to be recovered from any "W" of the "H" streams (Compl. ¶188).
  • Asserted Claims: Independent Claim 1 and dependent Claim 2 (Compl. ¶188).
  • Accused Features: Amazon S3 and Hadoop on EMR are accused of using this method through their "erasure coding" functionality, which allegedly divides data into "W" inputs (data shards), generates "H" output streams (data and parity shards), and applies checksums (Compl. ¶¶189, 195). The complaint includes a visual from an Amazon presentation illustrating the concept of "Data shards" and "Parity shards" to support this allegation (Compl. p. 107).

Multi-Patent Capsule: U.S. Patent No. 8,312,356 - *"Systems, Methods and Computer Program Products Including Features for Coding and/or Recovering Data"*

  • Patent Identification: U.S. Patent No. 8,312,356, "Systems, Methods and Computer Program Products Including Features for Coding and/or Recovering Data," issued on November 13, 2012.
  • Technology Synopsis: As a sibling to the ’944 patent, this patent addresses the same data durability challenge (Compl. ¶28). It claims a system embodied on computer-readable media for performing the erasure coding process, where the system includes a transformation component that processes "W" data inputs into "H" output streams using Galois field operations and involving the creation of H-sized intermediaries, and distributes the output streams between processing components (Compl. ¶219).
  • Asserted Claims: Independent Claim 17 and dependent Claim 20 (Compl. ¶219).
  • Accused Features: Amazon S3 and Hadoop on EMR are accused of being systems that operate on non-transitory media and include transformation components that perform the claimed erasure coding method (Compl. ¶¶220, 227).

III. The Accused Instrumentality

Product Identification

  • The complaint collectively identifies the accused instrumentalities as the "Transaction-Based Storage Patents Accused Products" and the "Data Coding Patents Accused Products" (Compl. ¶¶40, 54). Specific products named include Amazon Simple Storage Service (S3), Hadoop Distributed File System (HDFS) as used on Amazon Elastic MapReduce (EMR), the XFS file system as used on Amazon Elastic Compute Cloud (EC2) instances with Amazon Elastic Block Store (EBS), and the Lustre file system as used on Amazon FSx for Lustre (Compl. ¶40).

Functionality and Market Context

  • The accused products are core components of Amazon's cloud computing platform, providing scalable, durable, and high-performance data storage and processing (Compl. ¶¶2-3).
  • Amazon S3 is described as an object-based storage service that allegedly implements transaction logging via its "versioning" feature, which maintains a log of every object version (Compl. ¶42). It is also alleged to employ "striping techniques" by dividing and distributing data objects across arrays of servers to enhance performance (Compl. ¶43). A system architecture diagram from a presentation is provided to illustrate how S3 components are organized, showing data stored on groups of servers labeled "R2D2" (Compl. p. 27). For durability, S3 is alleged to use "erasure coding" to divide objects into data and parity shards stored across multiple drives (Compl. ¶55).
  • Hadoop on EMR is alleged to use HDFS, which implements a "write-ahead log" for transaction logging and employs striping by splitting files into blocks written across multiple "DataNodes" (Compl. ¶¶46-47). It is also alleged to use erasure coding for fault-tolerance (Compl. ¶57).

IV. Analysis of Infringement Allegations

'663 Patent Infringement Allegations

Claim Element (from Independent Claim 1) Alleged Infringing Functionality Complaint Citation Patent Citation
A transaction-logging data storage system comprising: a data storage subsystem that stores variable-size data objects... Amazon S3 is an object data storage system that is alleged to implement transaction logging via its "versioning" feature and stores objects ranging from 0 bytes to 5 TB. ¶¶63, 65 col. 4:51-54
...one or more storage devices comprising: an array of data storage units, wherein the array is configured to provide one or more ranks, each of the one or more ranks providing stripes for storing data objects... Amazon S3 allegedly stores data objects across arrays of servers organized into groups called "R2D2s" (ranks), transforming the objects into "shards" (stripes) for storage. ¶66 col. 6:2-6
...a rank manager that performs configuration processing associated with the one or more ranks... A combination of specialized servers, allegedly named C3PO, Skynet, Skywalker, and others, are said to manage the R2D2 server groups (ranks), including configuration tasks like adding new servers or removing faulty ones. ¶67 col. 6:7-13
...the data storage subsystem is configured to write data corresponding to each of the stripes into the one or more ranks as a unit... Amazon S3 is alleged to write and read data corresponding to shards (stripes) into the R2D2s (ranks) as a unit. ¶66 col. 5:14-16
a stripe space table that keeps track of space usage in the data storage subsystem... C3PO servers are alleged to communicate with R2D2 servers to determine available storage space and to track whether data stored in stripes is valid or no longer in use. ¶68 col. 9:30-36
a mapping component that processes mapping information between data object identifiers and physical address information of the data objects... The complaint alleges that Amazon S3's "keymap" and/or "Volume index" components process mapping information between object keys and the server addresses of the R2D2s where the data is stored. ¶69 col. 9:37-41
at least one processing component that performs operations among the data storage subsystem, the mapping component and the stripe space table, wherein the operations are coordinated to implement logging data objects... Skynet servers are alleged to coordinate operations by identifying available R2D2 servers (from the stripe space table function) and updating the key-to-server mappings (the mapping component function) to log the storage of data objects. ¶70 col. 5:1-13
...the mapping component includes one or more subcomponents that enable the transaction-logging data storage system to emulate one or more common data storage abstractions. The complaint alleges that the object key-to-server mappings can emulate a key-value store, and that because object keys can resemble file paths, the system can also emulate a traditional hierarchical folder structure. ¶71 col. 19:1-8
  • Identified Points of Contention:
    • Scope Questions: A central question may be whether the term "rank", which the patent describes in the context of disk arrays (e.g., RAID), can be construed to read on Amazon's alleged "groups of servers called R2D2s" (Compl. ¶66), which may be a more loosely-coupled, geographically distributed architecture. Similarly, the court may need to determine if Amazon's internally-named server functions like "C3PO" and "Skynet" perform the specific functions of the claimed "rank manager" and "stripe space table". A data flow diagram included in the complaint illustrates the alleged interaction between these named components (Compl. p. 29).
    • Technical Questions: The infringement analysis may focus on whether Amazon S3's "sharding" process is technically equivalent to the claimed "striping" process where data is written "as a unit." The complaint asserts this equivalence, but the actual implementation details of S3's data distribution will be a key factual issue.

V. Key Claim Terms for Construction

  • The Term: "rank"

  • Context and Importance: This term is foundational to the claimed system architecture. The complaint maps this term to "groups of servers called R2D2s" in the accused S3 system (Compl. ¶66). The construction of "rank" will be critical in determining whether Amazon's distributed, cloud-scale server architecture falls within the scope of claims rooted in descriptions of more traditional disk arrays.

  • Intrinsic Evidence for Interpretation:

    • Evidence for a Broader Interpretation: The specification introduces the term broadly, stating, "The present invention collects disk drives into groups called ranks" ('663 Patent, col. 9:60-61). This general description does not inherently limit a rank to a physically co-located or tightly-coupled set of disks.
    • Evidence for a Narrower Interpretation: The patent frequently discusses ranks in the context of RAID configurations and describes a "rank manager" that can reconfigure a rank by adding or removing individual disks ('663 Patent, col. 10:2-4). Figure 6 depicts ranks as distinct collections of physical disks, which might suggest a more localized and hardware-centric grouping than a logical collection of distributed cloud servers.

  • The Term: "transaction-logging"

  • Context and Importance: This term defines the fundamental data management approach of the invention. The complaint alleges that Amazon S3's "versioning" feature, which "maintains a persistent log of every version of each stored object," constitutes "transaction logging" (Compl. ¶¶42, 63). The construction will determine if S3's method of handling object updates meets this claim limitation. A screenshot from AWS documentation explaining "How S3 Versioning works" is included to support the logging allegation (Compl. p. 26).

  • Intrinsic Evidence for Interpretation:

    • Evidence for a Broader Interpretation: The patent specification contrasts the invention with "block oriented file systems" that overwrite data in place ('663 Patent, col. 4:4-8). It describes transaction logging systems as taking a "different approach" where the "entire filesystem can be viewed as a log, and new data and/or metadata can be written sequentially wherever there is available data storage" ('663 Patent, col. 4:44-48, via Compl. ¶23). This focuses on the non-destructive, append-only nature of writes.
    • Evidence for a Narrower Interpretation: The description of writing data "sequentially" could be a point of dispute. If construed to mean writing to physically contiguous blocks on a disk, it might not describe how a distributed object storage system places data. However, if construed more broadly as appending to a logical log rather than overwriting, it could encompass S3's alleged behavior of writing changes to "new storage space" (Compl. ¶65).

VI. Other Allegations

  • Indirect Infringement: The complaint alleges that Amazon induced infringement by marketing the accused products and providing documentation, user guides, and technical assistance that instruct and encourage customers to use the infringing functionalities (Compl. ¶99).
  • Willful Infringement: Willfulness is alleged based on pre-suit knowledge of the patents-in-suit. The complaint asserts that during the prosecution of its own patent application (which issued as the ’787 patent), Amazon was made aware of the publication for the ’528 patent when it was cited as prior art by the USPTO examiner. The complaint alleges Amazon "studied the specification for the Transaction-Based Storage Patents in detail" in its response to the examiner, long before the current lawsuit was filed (Compl. ¶¶89-90, 93).

VII. Analyst’s Conclusion: Key Questions for the Case

  • A core issue will be one of definitional scope: can terms like "rank" and "stripe", which are described in the patent specification with reference to RAID disk arrays, be construed to encompass the distributed, server-based architecture of a modern cloud storage service like Amazon S3, with its "groups of servers" and "shards"?
  • A key evidentiary question will be one of functional correspondence: what evidence will demonstrate that Amazon's proprietary, internally-named software components ("C3PO," "Skynet," "R2D2s") perform the specific, discrete functions of the "rank manager", "stripe space table", and "mapping component" recited in the claims, or is there a fundamental mismatch in their technical operation?
  • A central question for willfulness and potential damages enhancement will be one of pre-suit knowledge: does Amazon's interaction with the asserted patent family's specification during its own patent prosecution, as alleged in the complaint, establish knowledge of infringement sufficient to meet the standard for willful conduct?