Network-based computer system with improved network scheduling system

• US 5,414,845 A
• Filed: 06/26/1992
• Issued: 05/09/1995
• Est. Priority Date: 06/26/1992
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Eisenhauer, Daniel G., Balakrishnan, Govind, Behm, Jason L.
• Primary Examiner(s): Heckler, Thomas M.
• Application Number:
  US07/904,760
• Time in Patent Office: 1,047 Days
• Field of Search: 395/650, 395/700, 364/DIG. 1 MS File
• US Class Current: 718/104
• CPC Class Codes: G06F 11/20 using active fault-masking, e.g. by switching out faulty elements or by switching in spare elements
• CPC Class Codes: G06F 9/5044 considering hardware capabilities
• CPC Class Codes: G06F 9/5055 considering software capabilities, i.e. software resources associated or available to the machine
• CPC Class Codes: H04L 41/0894 Policy-based network configuration management
• CPC Class Codes: H04L 43/0805 by checking availability
• CPC Class Codes: H04L 43/0876 Network utilisation, e.g. volume of load or congestion level
• CPC Class Codes: H04L 9/40 Network security protocols

Abstract

An improved resource management system for a network-based computer system is described. The computer system includes a plurality of processors interconnected by a network where some of the processors are user nodes and others are batch nodes. The management system includes a delivery system for receiving the requests from the user nodes and a separate scheduler system for scheduling which request Go process next and on which batch node.

Claims

1. A network-based computer system comprising:

a network,

a plurality of processors coupled to said network and thereby to each other with some of said processors operating as user nodes for sending jobs to be run at batch nodes and some of said processors being the batch nodes and some being both batch and user nodes,

a resource management system for said computer system coupled to said network, said resource management system comprising;

a common scheduler for the computer system scheduling all jobs from user nodes to said batch nodes, said common scheduler including a policy module, said policy module for setting scheduling policy for said common scheduler for deciding the routing of jobs to batch nodes independently of the operation of said common scheduler so that said policy module can be replaced while the computer system is in operation to change said scheduling policy;

a delivery system responsive to jobs from said user nodes for sending scheduling requests to said common scheduler and in response to the decision of the scheduler, routing each job to one of said batch nodes said delivery system and said common scheduler being independent entities so that job delivery is separate from job scheduling, and

said common scheduler having a database storing job status and node availability information to run a job at a batch node, said common scheduler being responsive to said scheduling requests made through an application programming interface (API) from said delivery system and policy from said policy module for determining which job is scheduled next and on which node it is run.

2. The computer system of claim 1 wherein said common scheduler includes a master scheduler and at least one duplicate or spare scheduler for backing up said master scheduler, said spare scheduler maintaining a full copy of what is stored in said master scheduler for replacement of said master scheduler if said master scheduler is unavailable.

3. The computer system of claim 2 wherein said master and spare schedulers each have their own database and include their own policy module.

4. The computer system of claim 1 wherein said scheduler is responsive to a new job entering the system, a job starting or ending on a node, the availability of a node changing significantly or an operator action to alter the job which is scheduled next and on which batch processing node.

5. The computer system of claim 1 wherein said scheduler calls up said policy module each time a new job is received.

6. A resource management system for use in a computer system comprising a network, a plurality of processors coupled to said network and thereby to each other where some of said processors are user nodes for sending jobs and some are batch processors and some are both user nodes and batch processors to which jobs are sent, said resource management system comprising:

a common scheduler handling all requests from all user nodes in the system, said scheduler including a separate operationally independent policy module for setting forth rules for said scheduler so that the policy module can be replaced in a running system,

a delivery system responsive to jobs from said user nodes for sending scheduling requests to said common scheduler and in response to a routing message from said scheduler routing the job to the indicated batch processor, said delivery system being an independent entity from said scheduler so that job delivery is separate from job scheduling,

said scheduler having a database storing job status and node availability information to run a job at a batch node, said scheduling being responsive to said scheduling requests from said delivery system and policy from said policy module for determining which job is scheduled next and by which node and sending said routing message to said delivery system.

7. The computer system of claim 6 wherein said scheduler includes a master scheduler and at least one duplicate or spare scheduler for backing up said master scheduler if it should fail and having a full copy of what is stored in said master scheduler.

8. The computer system of claim 7 wherein said master scheduler and spare schedulers each have their own database and include their own policy module.

9. The computer system of claim 6 wherein said scheduler is responsive to a new job entering the system, a job starting or ending on a node, the availability of a node changing significantly or an operator action to alter the job which is scheduled next and on which batch processing node.

10. The computer system of claim 6 wherein said scheduler calls up said policy module each time a new job is received.

11. A method of scheduling jobs between user nodes and batch nodes in a computer system comprising a network and plural processors coupled to said network with same processors being at user nodes and other processors being at batch nodes of said network, jobs in said network being sent from said user nodes by means of user node requests and routed to said batch nodes by scheduling requests in accordance with a policy, and each batch node having availability and job status information, said method comprising the steps of:

sending said user node requests from said user nodes to a delivery system,

sending said scheduling requests from said delivery system to a common scheduler, said common scheduler operating independently of the delivery system and requesting access to a batch node in accordance with said policy,

sending said availability information and said job status information from said batch nodes to said common scheduler,

scheduling said jobs on said common scheduler based on said user node requests, said availability information, said job status information and said policy to provide which job is sent next and to what node, and

delivering said jobs using the delivery system to the batch nodes based on said scheduling whereby delivery and scheduling are maintained independent of one another.

12. The method set forth in claim 11 including the step of selectively changing the setting of said policy for the common scheduler independent of the operation of said delivery system so that policy for the scheduler can be changed in an operating computer system.