Method and apparatus for policy based class service and adaptive service level management within the context of an internet and intranet

• US 20040162901A1
• Filed: 02/19/2004
• Published: 08/19/2004
• Est. Priority Date: 12/01/1998
• Status: Active Grant

• Current Assignee: Krishna Mangipudi, Vijay Basani
• Original Assignee: Krishna Mangipudi, Vijay Basani
• Inventors: Mangipudi, Krishna, Basani, Vijay
• Granted Patent: US 7,124,188 B2
• Time in Patent Office: Days
• US Class Current: 709/225
• CPC Class Codes: H04L 41/0213 Standardised network management protocols, e.g. simple network management protocol [SNMP]
• CPC Class Codes: H04L 41/0604 using filtering, e.g. reduction of information by using priority, element types, position or time
• CPC Class Codes: H04L 41/0893 Assignment of logical groups to network elements
• CPC Class Codes: H04L 41/0894 Policy-based network configuration management
• CPC Class Codes: H04L 41/22 comprising specially adapted graphical user interfaces [GUI]
• CPC Class Codes: H04L 41/5009 Determining service level performance parameters or violations of service level contracts, e.g. violations of agreed response time or mean time between failures [MTBF]
• CPC Class Codes: H04L 41/5025 by proactively reacting to service quality change, e.g. by reconfiguration after service quality degradation or upgrade
• CPC Class Codes: H04L 43/045 for graphical visualisation of monitoring data
• CPC Class Codes: H04L 43/06 Generation of reports
• CPC Class Codes: H04L 43/08 Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
• CPC Class Codes: H04L 43/0811 by checking connectivity
• CPC Class Codes: H04L 43/0817 by checking functioning
• CPC Class Codes: H04L 43/0823 Errors, e.g. transmission errors
• CPC Class Codes: H04L 43/16 Threshold monitoring
• CPC Class Codes: H04L 67/02 based on web technology, e.g. hypertext transfer protocol [HTTP]
• CPC Class Codes: H04L 67/1001 for accessing one among a plurality of replicated servers
• CPC Class Codes: H04L 67/1006 with static server selection, e.g. the same server being selected for a specific client
• CPC Class Codes: H04L 67/1008 based on parameters of servers, e.g. available memory or workload monitoring of computer activity G06F11/30
• CPC Class Codes: H04L 67/1012 based on compliance of requirements or conditions with available server resources
• CPC Class Codes: H04L 67/1014 based on the content of a request
• CPC Class Codes: H04L 67/1017 : based on a round robin mechanism
• CPC Class Codes: H04L 67/1023 : based on a hash applied to IP addresses or costs
• CPC Class Codes: H04L 69/329 : in the application layer [OSI layer 7]

Abstract

A method and apparatus for robustly enhanced Class of Service (COS) at the application layer permits highly flexible privilege based access and enables implementation of complex policies and rules for classification and differentiation of services. Differentiation facilitates categorization of traffic to permit flexible design and implementation of multiple Class of Service levels. A routing host is configured to receive all client requests for sites and virtual sites implemented on a plurality of service hosts or back-end servers. A monitoring processor incorporating an Adaptive Policy Engine, in communication with the router (and agents on back-end servers) dynamically monitors workload and availability of servers to enable requests to be sent to the most appropriate and optimal server. Incoming traffic is first processed to assign a class. The APE is employed to monitor the incoming traffic to the routing host. Traffic is measured to each hosted site and further, to each class of a hosted site. The APE has a rules based engine that correlates this information and uses it to come up with a dynamic, real time balancing scheme for each hosted site. The APE or policy engine in conjunction with the router then intelligently distributes incoming traffic to the most available and/or efficient server within each class or “cluster,” by using one or more of a plurality of selectable load distribution algorithms, so that service level commitments are met. Intelligent agents deployed on each of the back-end servers monitor several server attributes/parameters and report back to the policy engine at the router. Class of service (COS) involves the classification of incoming requests by the policy engine. Backend server sites are clustered into virtual user definable cluster groups. Each cluster group can be managed/designated with a particular class of service. Based on information/parameters received, the composition of the clusters can be changed dynamically so that SLA parameters or metrics are within an acceptable range. Based on its class, the connection/request will be directed to one of the clusters. The specific machine selected will depend upon the load balancing algorithm defined for the cluster or class, and implemented as a function of the parameters reported to the policy engine, for making load balancing decisions.

Claims

1. A method for implementing class of service among a plurality of clients sending requests seeking access to sites hosted on a plurality of back-end servers, comprising the steps of:

grouping at least one of said plurality of back-end servers into a respective one of a plurality of service classes;

receiving a client request for host access at a front end processor;

selecting a class of service from said plurality of service classes according to at least one selected parameter of said request; and

distributing said request to a back-end server in said selected class of service according to the load of each of said at least one of said back-end servers in the selected service class.

2. The method of claim 1 in which said selected parameters of the request are selected from a group consisting of:

user authentication, virtual site level class of service and client level class of service;

wherein a user authentication identifies a subscribed class for an authenticated user;

a virtual site level class of service is determined by host name and selected protocol; and

a client level class of service is determined as a function of the request/transaction, service/protocol, authenticated user, URL, destination port, domain of origin, source IP, destination IP, and application requested.

3. The method of claim 1 in which said step of distributing the request according to the load further includes a load balancing algorithm selected from the group consisting of:

weighted percentage;

round robin;

CPU availability;

least connections; and

probabilistic.

4. A method for implementing probabilistic load balancing among a plurality of back-end servers, comprising the steps of:

determining the load on each of said plurality of back-end servers, including a maximum load L;

calculating the difference between the maximum load L+1 and the load of each of said plurality of back-end servers to obtain a respective value for each back-end server;

summing said respective values to obtain a value D; and

proportionally distributing the next D requests to the plurality of back-end servers based on their respective values such that more requests are routed to servers having a relatively light load.

5. A method for implementing a rules based adaptive policy engine for real-time balancing of incoming requests across a plurality of back-end servers comprising the steps of:

clustering said plurality of back-end servers according to class of service, each cluster having a corresponding class of service;

receiving a request characterized as belonging to a designated class of service;

checking to see if said request is a new session or an old session, where said old session has an identified back-end server and contents;

if said request is an old session, then

checking to see if said identified back-end server and contents are available;

if said request is a new session, or if said identified back-end server and content are not available,

running a load balancing algorithm selected for the designated class of service; and

routing said request to a back-end server within said cluster having the designated class of service, as selected by the load balancing algorithm.

6. The method of claim 5 in which said load balancing algorithm is selected from the group consisting of:

weighted percentage;

round robin;

CPU availability;

least connections; and

probabilistic.

7. The method of claim 5 in which the step of running a load balancing algorithm further comprises the steps of:

deploying an intelligent agent on at least one of said plurality of back-end servers;

using said intelligent agent to collect information in the form of selected server attributes;

transmitting said collected information to an adaptive policy engine; and

selecting a back-end server according to the collected server attributes.

8. The method of claim 7 in which said step of transmitting said collected information further includes:

a network transmission step of multicasting a UDP packet containing information related to said selected server attributes; and

a network reception step of decoding the UDP packet in the adaptive policy engine.

9. A method for implementing collection of information for monitoring the performance in a cluster of web servers, comprising the steps of:

providing an adaptive policy engine;

deploying an intelligent agent on at least one of said plurality of web servers;

using said intelligent agent to collect information in the form of selected server attributes;

transmitting said collected information to said adaptive policy engine for use in dynamically allocate requests to selected web servers in said cluster of web servers to meet pre-defined Service Level Agreements (SLA).

10. The method of claim 9 further including the step of repackaging at least some of said collected information into a management information base.

11. The method of claim 9 further including the step of making at least some of said collected information available to an application program for monitoring real-time web server performance.