Obviousness

Obviousness Analysis of US9215275 under 35 U.S.C. § 103

This analysis evaluates whether the claims of US Patent 9215275, "System and method to balance servers based on server load status," would have been obvious to a person having ordinary skill in the art (POSITA) at the time of the invention (priority date: September 30, 2010). The analysis considers the prior art references explicitly cited within the patent.

Level of Ordinary Skill in the Art

A person having ordinary skill in the art in the field of US Patent 9215275 would be a network engineer or software developer with experience in data communications, server load balancing, distributed systems, and web services. Such an individual would be familiar with common network protocols (e.g., HTTP, FTP, SIP, XML), server architectures (e.g., web servers, application servers), and techniques for monitoring server health and distributing network traffic.

Differences Between Prior Art and Claims

The independent claims (Claim 1, 6, and 11) of US9215275 generally describe a method, computer-readable medium, and system for a service gateway to balance servers by:

1. Receiving service responses from servers, where these responses include a server status indicating the computing load status of the server.
2. Obtaining this server status from the service response.
3. Receiving a new service request with a URL.
4. Determining primary/secondary server configurations for the URL.
5. Discovering, using the server status from the service response, that a primary server temporally cannot act as primary.
6. Discovering, using the server status from the service response, that a secondary server temporally can act as secondary.
7. Directing the new service request to the available secondary server.

As noted in the prior art section, a key distinguishing feature of US9215275 is the explicit disclosure of obtaining server computing load status directly from a service response to dynamically adjust server selection for new service requests. The cited prior art references generally address aspects of load balancing or server monitoring but do not explicitly combine these elements in the same manner.

Specifically:

• US20020194350A1 (Lu Leonard L.) teaches a content-aware web switch that directs requests based on URLs, which aligns with the service gateway's function of receiving requests and routing based on URL association. However, it does not explicitly disclose servers sending their real-time computing load status within service responses.
• US20040202182A1 (Lund Martin) introduces the concept of load balancing using server parameters, specifically "spare link bandwidth" in a blade server environment. This reference anticipates the general idea of using server status to determine if a server "temporally cannot act". However, it focuses on a network-level metric (bandwidth) rather than application-level computing load status, and does not specify obtaining this status within a service response.
• US7076555B1 (Novell, Inc.), US20070195792A1 (A10 Networks Inc.), and US20100235880A1 (A10 Networks, Inc.) address TCP connection takeover, SYN cookie mechanisms, and network traffic policies, respectively. While relevant to network and server management, their inventive concepts are distinct from the specific mechanism of server load balancing based on computing load status embedded in service responses.

Motivation to Combine Prior Art References

A POSITA at the time of the invention would have been motivated to combine the teachings of the prior art and modify them to arrive at the claimed invention due to a recognized problem in existing load balancing systems. The background section of US9215275 explicitly articulates this problem: "Common methods to balance load among servers is to distribute the service requests based on the applications (HTTP, FTP, etc.), service addresses such as URL, priorities based on network interfaces or host IP addresses. SLB may distribute service requests additionally in a round robin fashion to the servers, assuming and ensuring the servers would be evenly loaded. However, different service requests have different service computing consequences. A server may be fully loaded with only a handful of service requests while another server remains mostly idle even with plenty of service requests. SLB may inappropriately send another request to a busy server, incorrectly considering the busy server being readily available, instead of sending the request to an idle server."

This statement clearly highlights the known deficiency of traditional load balancers in accurately assessing real-time server computing load status to make intelligent distribution decisions.

The motivation to combine and modify would be to enhance the accuracy and responsiveness of load balancing by providing the service gateway with a more direct and real-time measure of a server's actual computing capacity.

1. Combining Content-Aware Routing with Server Load Awareness: A POSITA would logically combine the content-aware routing capabilities of a web switch (as taught by US20020194350A1) with the general principle of using server load information for selection (as taught by US20040202182A1). While US20040202182A1 uses "spare link bandwidth," a POSITA would recognize that application-level services require knowledge of internal computing resources (CPU, memory, storage utilization, etc.) for optimal load distribution, as explicitly stated in US9215275.
2. Modifying Server Status Reporting to Use Service Responses: Given that the service gateway (load balancer) already intercepts and processes service responses (as part of its function described in US20020194350A1), it would be an obvious design choice for a POSITA to embed the server's computing load status directly within these existing communication flows. This approach is efficient, avoids the overhead of separate out-of-band monitoring or polling mechanisms, and provides context-specific load information. The patent itself suggests various common methods for embedding this status, such as in an HTTP header, HTML tag, XML tag, SIP tag, or FTP reply code. Extending existing protocols with custom tags or headers is a well-known and routine engineering practice for conveying additional metadata.
3. Applying Primary/Secondary Server Logic with Real-Time Status: The concept of configuring servers as primary or secondary for specific URLs is a common practice in load balancing for fault tolerance and optimized resource allocation, and is explicitly described in US9215275 as a known configuration. Once a more accurate, real-time computing load status is available (via embedding in service responses), it would be an obvious and natural extension for the service gateway to use this enhanced information to intelligently route traffic. If a primary server reports it is overloaded (e.g., status '2' for "very busy" as shown in FIG. 3 and discussed in the patent), a POSITA would logically redirect new requests for that URL to an available secondary server to prevent service degradation, thus addressing the problem identified in the patent's background.

Therefore, motivated by the known problem of inefficient server load balancing due to a lack of accurate, real-time computing load status, a POSITA would have been led to combine the content-aware routing of US20020194350A1 with the concept of using server-reported load from US20040202182A1, and to implement the server load reporting by embedding computing load status in application-level service responses using conventional protocol extension techniques. This combination and modification would have been made with a reasonable expectation of improving load balancing efficiency and addressing the identified problem.