Cluster arrangement
First Claim
1. A cluster arrangement of a system, comprising:
- a first network;
a plurality of data processing systems, each of said plural data processing systems forming an application node and having an operating system which is the operating system in each of said plural data processing systems, each said application node including an entity which is executed;
an inspection node connected to the first network, the inspection node including the operating system and an inspection agent configured to manage reports received from each application node; and
an application agent on the each application node, the application node comprising a service of the operating system and being started following initialization of the operating system on the each application node, the application agent being configured to;
monitor and identify entities executed on the each application node;
start a new entity or restart an executed entity which has been terminated prematurely on the each application node;
assess and decide whether the new or restarted entity can be executed on the each application node;
forward a request to execute the new or restarted entity to the application agents at all of the application nodes connected to the first network;
report to the application agents at all of the application nodes connected to the network following an acceptance of the request to execute the new or restarted entity; and
report to the application agents at all of the application nodes connected to the network if an application agent rejects the request to execute the new or restarted entity.
3 Assignments
0 Petitions
Accused Products
Abstract
A cluster arrangement having a first network, at least two data processing systems which each form an application node and which each have an operating system, and an application agent on each application node, which application agent is operable to monitor and identify entities executed on the application node; start a new entity or restart an entity which has been terminated prematurely on the application node; assess and decide whether it is possible to execute a new entity on the application node; send a request for execution of an entity to application agents at application nodes connected to the network; and report to application agents at application nodes connected to the network following an acceptance of the request for execution of an entity.
-
Citations
34 Claims
-
1. A cluster arrangement of a system, comprising:
-
a first network; a plurality of data processing systems, each of said plural data processing systems forming an application node and having an operating system which is the operating system in each of said plural data processing systems, each said application node including an entity which is executed; an inspection node connected to the first network, the inspection node including the operating system and an inspection agent configured to manage reports received from each application node; and an application agent on the each application node, the application node comprising a service of the operating system and being started following initialization of the operating system on the each application node, the application agent being configured to; monitor and identify entities executed on the each application node; start a new entity or restart an executed entity which has been terminated prematurely on the each application node; assess and decide whether the new or restarted entity can be executed on the each application node; forward a request to execute the new or restarted entity to the application agents at all of the application nodes connected to the first network; report to the application agents at all of the application nodes connected to the network following an acceptance of the request to execute the new or restarted entity; and report to the application agents at all of the application nodes connected to the network if an application agent rejects the request to execute the new or restarted entity. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26)
-
-
27. A method in a cluster arrangement comprising a plurality of application nodes and an inspection node including a device for controlling an initialization operation at each of said plurality of application nodes, each of said plurality of application nodes and the inspection node being connected to a network, the method comprising the steps of:
-
receiving at the inspection node from each of said plurality of application nodes a first list including all entities executed on the each of said plurality of application nodes and data required to execute the entities, the initialization operation including a command for loading the operating system of the application node to each of said plurality of application nodes; checking with the inspection node, at regular intervals of time, to determine whether any of said plurality of application nodes has failed; if any of said plurality of application nodes has failed, compiling with the inspection node a second list of entities executed on the failed application node and the data required to execute the entities executed on the failed application node, and forwarding the second list to all of said plurality of application nodes connected to the network with a request to execute the entities on the forwarded second list; assessing and deciding, at each of said plurality of application nodes, whether to start a new entity or restart a failed entity upon receipt of the request to execute an entity on the forwarded second list; and reporting to application agents at all of said plurality of application nodes if an application agent rejects the request to execute the entity on the forwarded second list. - View Dependent Claims (28, 29, 30, 31, 32, 33, 34)
-
Specification