Google LLC v. Sandpiper CDN LLC

1. Case Identification

• Case #: IPR2025-00806
• Patent #: 8,645,517
• Filed: April 14, 2025
• Petitioner(s): Google LLC
• Patent Owner(s): Sandpiper CDN, LLC
• Challenged Claims: 1-20

2. Patent Overview

• Title: Policy-Based Domain Name Service
• Brief Description: The ’517 patent discloses a system and method for managing network traffic in a content delivery network (CDN). It describes using a graphical user interface (GUI) to create and implement traffic-directing policies, which are structured as a hierarchical decision tree, for a domain name service (DNS).

3. Grounds for Unpatentability

Ground 1: Obviousness over Beaumont and Hasan - Claims 1, 4-9, and 12-20 are obvious over Beaumont in view of Hasan.

• Prior Art Relied Upon: Beaumont (Application # 2002/0169890) and Hasan (Patent 7,082,464).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner argued that Beaumont disclosed a CDN that uses policy-based DNS to direct traffic. Beaumont’s system selected content servers based on customizable preference values stored in a database, which could be structured as a tree. However, Beaumont did not explicitly disclose a GUI for setting these preferences. Hasan disclosed a network management system with a Java-based GUI that used a “familiar tree structure” to allow administrators to configure device and policy information. The combination, Petitioner asserted, taught a method of providing Hasan's tree-based GUI to set the traffic-directing policies in Beaumont's CDN framework. The branches of Hasan’s tree map to Beaumont’s decision criteria (e.g., server health, client domain), and the leaves map to Beaumont’s content servers (the "answers" to a DNS request).
  • Motivation to Combine: A person of ordinary skill in the art (POSITA) would combine the references to provide an intuitive, user-friendly interface for managing Beaumont's manually entered preference rules. Using a known GUI like Hasan’s to improve the usability of a similar system like Beaumont's was a predictable design choice.
  • Expectation of Success: The combination was straightforward, as Beaumont already suggested a tree hierarchy for its preference database, and Hasan provided a well-understood, platform-independent GUI for implementing such a structure.

Ground 2: Obviousness over Beaumont, Hasan, and Biliris - Claims 2, 3, 10, and 11 are obvious over Beaumont and Hasan in view of Biliris.

• Prior Art Relied Upon: Beaumont (Application # 2002/0169890), Hasan (Patent 7,082,464), and Biliris (Application # 2002/0078233).
• Core Argument for this Ground:
  • Prior Art Mapping: This ground builds upon the Beaumont and Hasan combination to address claims requiring policies based on sub-domains (e.g., "images.xyz.com" as distinct from "www.xyz.com"). Petitioner contended that while the Beaumont/Hasan system provided policy-based traffic direction for a domain, Biliris explicitly taught a DNS outsourcing mechanism for redirecting queries based on specific sub-domains to improve load balancing and network efficiency.
  • Motivation to Combine: A POSITA would incorporate Biliris’s teaching of subdomain-level request distribution into the Beaumont/Hasan system to gain more granular control over traffic and achieve better load distribution. Biliris itself noted that its subdomain approach was an improvement over handling traffic for an entire domain, providing a clear reason for its adoption.
  • Expectation of Success: Since Biliris's system was analogous to Beaumont's CDN architecture, implementing subdomain delegation within the Beaumont/Hasan framework would have been a predictable integration with a high expectation of success.

Ground 3: Obviousness over Ben-Shaul and Hasan - Claims 1-20 are obvious over Ben-Shaul alone or in view of Hasan.

• Prior Art Relied Upon: Ben-Shaul (Application # 2002/0010798) and Hasan (Patent 7,082,464).
• Core Argument for this Ground:
  • Prior Art Mapping: Petitioner asserted that Ben-Shaul alone rendered all challenged claims obvious, as it taught a CDN that used policy-based DNS ("directives") to route traffic. Ben-Shaul disclosed multiple GUIs, including a "highly graphical" editor and a "front edge tool" with a "hierarchical view," for constructing these policies. This hierarchical grouping of edge servers (leaves) under logical nodes (branches) functioned as the claimed decision tree. In the alternative, Petitioner argued that combining Ben-Shaul with Hasan's "familiar tree structure" GUI further supported the obviousness of a decision-tree interface for setting Ben-Shaul's policies.
  • Motivation to Combine: A POSITA would be motivated to utilize Hasan’s well-known, platform-independent tree-structure GUI to implement Ben-Shaul's policy directives, which Ben-Shaul itself taught should be made easy to construct. This combination would represent the use of a known tool (Hasan's GUI) for its intended purpose to improve a similar system (Ben-Shaul's).
  • Expectation of Success: The combination would have been straightforward, as both references described using web-based GUIs for policy editing, and incorporating Hasan's tree structure into Ben-Shaul's system was a predictable design choice.

4. Key Technical Contentions (Beyond Claim Construction)

• Priority Date Challenge: Petitioner argued that the ’517 patent was not entitled to its claimed 2001 priority date. Key limitations recited in the independent claims, such as resource nodes being "mail exchange (MX) records," were first introduced in a 2007 continuation-in-part application. Therefore, the effective priority date is, at best, October 26, 2007, which makes all asserted prior art references, published between 2002 and 2006, available under §102.

5. Arguments Regarding Discretionary Denial

• §325(d) Denial Inappropriate: Petitioner argued that discretionary denial under §325(d) was improper because the asserted prior art and arguments were not previously considered by the USPTO. While some references were cited in an Information Disclosure Statement, there was no evidence the Examiner substantively reviewed them against the claims. Furthermore, the petition presented new combinations and the benefit of an expert declaration, creating a different record than what was before the Examiner.
• Fintiv Denial Inappropriate: Petitioner argued that the Fintiv factors weighed in favor of institution. The co-pending district court case was in its early stages, with fact discovery just beginning and a trial date (September 2026) that aligns closely with the anticipated Final Written Decision (FWD) date (October 2026). Petitioner also stipulated that it would not pursue the same grounds in district court if the IPR is instituted, mitigating concerns of duplicative efforts.

6. Relief Requested

• Petitioner requests institution of an inter partes review and cancellation of claims 1-20 of the ’517 patent as unpatentable.