Method and apparatus for scheduling jobs on a network

• US 7,831,972 B2
• Filed: 11/03/2005
• Issued: 11/09/2010
• Est. Priority Date: 11/03/2005
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Yasskin, Jeffrey, Wolf, Alyssa A., Zeigler, Benjamin J., Schwichtenberg, Frank, Moore, Matthew P., Redburn, Jeremy A.
• Primary Examiner(s): An; Meng-Ai
• Assistant Examiner(s): Lee; Adam
• Application Number:
  US11/266,804
• Publication Number:
  US 20070101336A1
• Time in Patent Office: 1,832 Days
• Field of Search: 718102-103, 717168-178, 705/59, 709/220
• US Class Current: 718/102
• CPC Class Codes: G06F 8/60 Software deployment
• CPC Class Codes: G06F 9/505 considering the load
• CPC Class Codes: G06F 9/5055 considering software capabilities, i.e. software resources associated or available to the machine

Abstract

An apparatus and method for scheduling a job process on at least one node in a server data processing network. The scheduling is based on a state of at least two nodes in the network as well as a cost metric of installing a required application on at least one of the nodes in the network of nodes. The apparatus and method gathers metrics indicating properties associated with at least two nodes on the network, the metrics including a presence of an application necessary for running the process. A first cost factor for scheduling the process on a node in the network having the application necessary for running the process is compared with a second cost factor for scheduling the process on a node in the network that does not have the application necessary for running the process. Finally, a node in the network is selected for scheduling the process.

Claims

1. A method for selecting at least one computational node in a distributed information processing system for scheduling a job process, the method on a node manager comprising:

receiving a job to be scheduled for execution on at least one of a plurality of computational nodes communicatively coupled together over a network;

determining that the job is dependent on at least one software application;

determining that a first computational node in the plurality of computational nodes comprises the software application;

determining that an availability of a set of resources necessary for executing the job on the first computational node is less than an availability of a set of resources on at least a second computational node, wherein the second computational node fails to comprise the software application;

determining a first computational cost associated with the first computational node based on the availability of set of resources for executing the job on the first computational node at the first computational cost;

determining a second computational cost associated with a second computation node based on installation metrics associated with installing the software application on the second computational node, wherein the installation metrics comprise at least installation time of the software application, disk, processor, and memory requirements of the software application, and a licensing cost of the software application;

comparing the first computational cost with the second computation cost; and

selecting between the first and second computational nodes the computational node with a lowest computational cost for executing the job.

2. The method according to claim 1, wherein the dependency indicates at least one of:

a required second application; and

a conflicting second application.

3. The method according to claim 1, further comprising:

installing the at least one application necessary for running the job in response to the selected node not having the at least one application necessary for running the job.

4. The method according to claim 1, wherein the metrics include at least one of:

a quantity of incoming processes;

a type of incoming processes; and

currently installed applications.

5. The method according to claim 1, wherein the computational cost includes at least one of:

a time required to install the necessary application; and

a licensing cost.

6. The method according to claim 5, wherein the time required to install the necessary application is specified by a user.

7. The method according to claim 5, wherein the time required to install the necessary application is based on at least one of:

an average time of past installs;

a median time of past installs; and

a highest time of past installs.

8. The method according to claim 1, wherein elements within the computational cost are assigned weights.

9. An apparatus for scheduling a job process on at least one computational node in a distributed information processing system, the apparatus comprising:

a node manager for receiving a process to be scheduled;

a plurality of computational nodes communicatively coupled to the node manager;

a storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for;

determining that the job is dependent on at least one software application;

determining that a first computational node in the plurality of computational nodes comprises the software application;

determining that an availability of a set of resources necessary for executing the job on the first computational node is less than an availability of a set of resources on at least a second computational node, wherein the second computational node fails to comprise the software application;

determining a first computational cost associated with the first computational node based on the availability of set of resources for executing the job on the first computational node at the first computational cost;

determining a second computational cost associated with a second computation node based on installation metrics associated with installing the software application on the second computational node, wherein the installation metrics comprise at least installation time of the software application, disk, processor, and memory requirements of the software application, and a licensing cost of the software application;

comparing the first computational cost with the second computation cost; and

a selector that selects, based on a comparison of the first computational cost and the second computational cost, the first node or the second node in the network for scheduling the process.

10. The apparatus according to claim 9, further comprising:

a provisioning system for provisioning onto the selected node the application necessary for running the process; and

a scheduler for scheduling the process on the selected computer.

11. The apparatus according to claim 9, wherein the metrics include at least one of:

quantity of incoming processes;

type of incoming processes; and

currently installed applications.

12. The method according to claim 9, wherein the computational cost includes at least one of:

a time required to install the necessary application;

a licensing cost; and

historical time taken to provision an application.

13. A computer readable storage product stored in a memory for selecting at least one computational node in a distributed information processing system for scheduling a job process, the computer readable storage product when executed by a processing circuit performing a method comprising:

receiving a job to be scheduled for execution on at least one of a plurality of computational nodes communicatively coupled together over a network;

determining that the job is dependent on at least one software application;

determining that a first computational node in the plurality of computational nodes comprises the software application;

determining that an availability of a set of resources necessary for executing the job on the first computational node is less than an availability of a set of resources on at least a second computational node, wherein the second computational node fails to comprise the software application;

determining a first computational cost associated with the first computational node based on the availability of set of resources for executing the job on the first computational node;

determining a second computational cost associated with the second computation node based on installation metrics associated with installing the software application on the second computational node, wherein the-installation metrics comprise at least installation time of the software application, disk, processor, and memory requirements of the software application, and a licensing cost of the software application;

comparing the first computational cost with the second computation cost; and

selecting between the first and second computational nodes the computational node with a lowest computational cost for executing the job.

14. The computer readable storage product according to claim 13, wherein determining that an availability of the set of resources necessary for executing the job on the first computational node is less than the availability of the set of resources on at least a second computational node is based on:

a quantity of incoming processes; and

a type of incoming processes.

15. The computer readable storage product according to claim 13, wherein the installation time of the software application is specified by a user.

16. The computer readable storage product according to claim 13, wherein the installation time of the software application is based on at least one of:

an average time of past installs;

a median time of past installs; and

a highest time of past installs.

17. The computer readable storage product according to claim 13, wherein elements within the first computational cost and the second computational cost are assigned weights.