Method and apparatus for cross-partition control in a partitioned process environment

• US 5,345,590 A
• Filed: 09/01/1993
• Issued: 09/06/1994
• Est. Priority Date: 08/31/1990
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Petersen, David B., Redding, Ian G., Ault, Donald F., Schmandt, Stephen J.
• Primary Examiner(s): Shaw, Gareth D.
• Assistant Examiner(s): KATBAB, ABDOLLAH
• Application Number:
  US08/115,913
• Time in Patent Office: 370 Days
• Field of Search: 395/650, 364/DIG. 1, 364/280.6, 364/281.9, 364/285.1, 364/285.3
• US Class Current: 718/1
• CPC Class Codes: G06F 11/0712 in a virtual computing platform, e.g. logically partitioned systems
• CPC Class Codes: G06F 11/0757 by exceeding a time limit, i.e. time-out, e.g. watchdogs
• CPC Class Codes: G06F 11/2025 using centralised failover control functionality
• CPC Class Codes: G06F 11/2035 without idle spare hardware
• CPC Class Codes: G06F 11/3466 Performance evaluation by tracing or monitoring
• CPC Class Codes: G06F 15/161 Computing infrastructure, e.g. computer clusters, blade chassis or hardware partitioning casings, cabinets, racks or drawers for data centers H05K5/00

Abstract

A logically partitioned data processing system has a policy defining responsive actions to be undertaken by a process in one partition because of a monitored failure of an operating system in another partition. When such a failure occurs, the monitoring partition, if authorized, automatically communicates with a hypervisor to initiate the responsive actions on the failing partitions to reset and/or reconfigure that failing partition. Communication of action request between the partition and the hypervisor is accomplished without operator intervention, through a Service Call Logical Processor interface.

Claims

1. In a logically partitioned data processing system comprising a plurality of physical processors partitioned into two or more logical partitions, each partition having a memory containing an associated control program and each partition capable of having a different status, apparatus for cross-partition control comprising:

a) authorization means connected to each said partition for authorizing a controlling process to be executed by a control program in any one of the two or more logical partitions to initiate responsive actions on one or more processors of another one of the two or more logical partitions;

b) policy definition means located in the memory of each said partition specifying a set of trigger events and associated responsive action definitions, each responsive action definition in the set specifying at least one of said responsive actions for an associated control program to perform when said trigger events occur;

c) monitor means, in said any one of the two or more logical partitions, for monitoring the status of others of the two or more logical partitions to detect the trigger event in one of said others of the two or more logical partitions;

d) service processor means connected to said partitions for initiating said responsive action to reconfigure said memory and processors of one of said another one of said two or more logical partitions, and to cause said control program in said any one of the two or more logical partitions to execute said controlling process to direct actions in the reconfigured memory and processors; and

e) communication means located in each said partition for communicating a request from said controlling process in said any one of the two or more logical partitions to said service processor means to initiate said responsive action when the monitor means has so detected said trigger event, and the authorization means has so authorized execution of said controlling process to initiate a responsive action specified by said trigger even in said policy definition means.

2. The apparatus of claim 1 wherein said service processor means includes said authorization means, and each set of responsive action definitions comprises a cross-partition deactivate means for requesting said service processor to deactivate and reset said another one of said partitions.

3. The apparatus of claim 1 wherein each set of responsive action definitions further comprises a resource reconfiguration means for requesting said service processor to reconfigure said memory and processors of said another one of said partitions.

4. The apparatus of claim 3 wherein the communication means comprises an SCCB transmission means for performing a Service Call Logical Processor function having an SCCB interface.

5. In a logically partitioned data processing system comprising a service processor and a plurality of physical processors partitioned into two or more logical partitions, each partition having a memory containing an associated control program for executing a controlling process and each capable of having a different status, a method for cross-partition control comprising the steps of:

a) authorizing said associated control program to execute a controlling process in any one of the two or more logical partitions to initiate responsive actions on one or more processors of another one of the two or more logical partitions;

b) loading policy-definition data in the memory of each of said logical partitions, each said policy-definition data comprising a set of trigger events and associated responsive action definitions, each responsive action definition specifying a responsive action for an associated control program to perform when said trigger events occur;

c) said any one of said logical partitions monitoring the status of others of the two or more logical partitions to detect the trigger event in one of said others of the two or more logical partitions;

d) initiating said responsive action via said policy-definition data to cause said service processor to reconfigure said memory and processors of said another one of said partitions, and to cause said controlling process in said any one of the two or more logical partitions to direct actions in said reconfigured memory and processors; and

e) communicating a request from said any one of said logical partitions to said service processor for said controlling process to perform the responsive action when one of the two or more logical partitions has so detected said trigger event, and said controlling process has been authorized, and said responsive action is specified by a responsive action definition.

6. The method of claim 5 wherein said sets of responsive action definitions perform a cross-partition deactivate action causing said service processor to deactivate and reset said another one of said partitions in response to a responsive action definition.

7. The method of claim 6 wherein the set of responsive action definitions further perform a resource reconfiguration action causing said service processor to reconfigure said memory and processors of said another one of said partitions in response to said responsive action definitions.

8. The method of claim 7 wherein the step of communicating includes performing an SCCB transmission by a Service Call Logical Processor function having an SCCB interface.