Managing a computer system with blades

• US 7,610,582 B2
• Filed: 03/25/2004
• Issued: 10/27/2009
• Est. Priority Date: 04/18/2003
• Status: Active Grant

• Current Assignee: SAP SE
• Original Assignee: SAP AG (SAP SE)
• Inventors: Von Bergen, Axel, Schwarz, Arne, Sauermann, Volker, Becker, Wolfgang, Zachmann, Guenter
• Primary Examiner(s): Zhen; Wei Y
• Assistant Examiner(s): Brophy; Matthew J
• Application Number:
  US10/553,607
• Publication Number:
  US 20070083861A1
• Time in Patent Office: 2,042 Days
• Field of Search: 717168-178
• US Class Current: 717/171
• CPC Class Codes: G06F 8/60 Software deployment

Abstract

Managing a computer system having a plurality of blades by detecting the presence of a new blade in the computer system; installing an operating system on the new blade; configuring the operation system; and copying a service that is running on an earlier detected blade to the new blade. Further features are testing services and operating system on the new blade and cyclically shifting services through the blades.

Claims

1. A computer-implemented method for managing a computer system, the computer system operating with a plurality of blades, the method comprising steps performed by a computer of:

detecting the presence of a new blade in the computer system, wherein the plurality of blades includes first and second groups, and the new blade belongs to the first group;

monitoring system performance;

upon reaching a predefined threshold of a measurement value, automatically installing a first service on the new blade and a second service to a second blade in the second group of blades, wherein the first and second services differ, wherein the measurement values are taken from at least the following;

usage of processor resources, processing times, usage of memory, remaining capacity of data storage, and communication parameters of a blade interface;

automatically configuring the first service based on a configuration used in an earlier detected blade;

copying a service that is running on the earlier detected blade to the new blade, wherein the service is copied from a memory of the earlier detected blade to the new blade;

testing the service in parallel operation on the earlier detected blade and the new blade;

cyclically repeating, in a first cycle, copying, installing, and parallel testing for the plurality of blades, wherein a first specialized configuration is used for the service of the first group;

cyclically repeating, in a second cycle, copying, installing, and parallel testing for the plurality of blades in the second group while keeping the number of blades in the second group that are installing the service smaller than the number of blades in the second group that are not installing the service, wherein a second specialized configuration is used for the service of the second group; and

cyclically repeating, in a super cycle across the first and second groups, copying, installing, and parallel testing of a third service across the first and second groups, where the number of blades installing the third service is kept less than the number of blades not installing the third service,

wherein the copying of the second service is monitored by a group of blades other than the second group, and the shifting of the first service is monitored by a group of blades other than the first group.

2. The method of claim 1, wherein installing the operating system is performed by accessing a mass storage that is part of the computer system.

3. The method of claim 1, wherein installing is performed by using scripts.

4. The method of claim 3, wherein installing is performed by using scripts that are part of the service that is running on the computer system prior to detecting the new blade.

5. The method of claim 1, wherein monitoring is performed periodically.

6. The method of claim 1, wherein monitoring is performed by monitoring processes that operate consecutively for adjacent blades.

7. The method of claim 6, wherein monitoring is performed by a token ring technique.

8. The method of claim 1, wherein the measurement values are related to the blades independently.

9. The method of claim 1, wherein the processing times are related to processing times for incoming telephone calls and a call rate.

10. The method of claim 1, wherein computer instructions to perform the detecting step are part of services that are running on the computer system.

11. The method of claim 1, wherein computer instructions for the detecting and copying steps are performed according to criteria in the service that is running on the earlier detected blade.

12. The method of claim 1, wherein copying the service comprises restarting the service, wherein executable instructions of the service are loaded from a central storage and wherein an image of the process context of the service is transferred to the new blade.

13. The method of claim 1, wherein copying the service comprises modifying the version of the service.

14. The method of claim 1, wherein installing the operating system comprises modifying the system.

15. The method of claim 1, wherein the method is performed for at least 3 blades, the method further comprising the subsequent execution of a controller service, an engine service, and a monitor service, the services belonging to a same business application.

16. The method of claim 1, controlled by a controller residing on at least one blade, wherein the controller further performs at least one function selected from the group of:

testing the copy of the service on the new blade, and modifying the execution of the service on the earlier detected blade if the copy of the service operates successfully.

17. The method of claim 16, wherein modifying comprises stopping the service on the earlier detected blade.

18. A processor-implemented method for managing a computer system, the system operating with a plurality of computers in at least two groups, the method comprising steps performed by a processor of:

assigning, via the processor, a first service to the first group of computers and a second service to the second group of computers, wherein the first and second services differ;

monitoring system performance;

upon reaching a predefined threshold of a measurement value, shifting, via the processor, the first service that runs on a first computer of the first group to run on a second computer in the first group, wherein the shifting includes copying the first service from a memory of the first computer to the second computer, wherein the first and second services differ, and wherein the measurement values are taken from at least;

usage of processor resources, processing times, usage of memory, remaining capacity of data storage, and communication parameters of a blade interface;

testing, via the processor, the service in parallel operation on the first computer and on the second computer, and disabling the operation of the service by the first computer only if the test is successful;

installing the service to the first computer;

cyclically repeating, in a first cycle, the shifting, re-installing, and parallel testing for the plurality of computers in the first group while keeping the number of computers in the first group that are re-installing the first service smaller than the number of computers that are not installing the first service, wherein a first specialized configuration is used for the first service of the first group;

cyclically repeating, in a second cycle, shifting, installing, and parallel testing for the plurality of computers in the second group while keeping the number of computers in the second group that are installing the second service smaller than the number of blades in the second group that are not installing the second service, wherein a second specialized configuration is used for the service of the second group; and

cyclically repeating, in a super cycle across the first and second groups, a shifting, installing, and parallel testing of a third service across the first and second groups, where the number of computers installing the third service is kept less than the number of blades not installing the third service,

wherein the shifting of the second service is monitored by a group of computers other than the second group, and the shifting of the first service is monitored by a group of computers other than the first group.

19. The method of claim 18, wherein shifting and re-installing is repeated cyclically for all computers in the groups, thereby keeping the number of computers that are re-installing the operating system smaller than the number of computers that are not re-installing the operating system.

20. The method of claim 18, wherein the assigning step is performed for services of a first class on a first group of computers and for services of a second class on a second group of computers.

21. The method of claim 18, wherein the computers are blades.

22. A tangible computer-readable storage medium comprising instructions for execution by a processor for the practice of a method for managing a computer system, the instructions, when executed by the processor, causing the processor to:

detect the presence of a new blade in the computer system, wherein the computer system includes a plurality of blades in at least first and second groups;

automatically install an operating system on the new blade;

automatically configure the operating system based on a configuration used in an earlier detected blade;

monitor system performance;

upon reaching a predefined threshold of a measurement value, copy a first service that is running on the earlier detected blade from the earlier detected blade to the new blade, wherein the first service applies to blades in the first group and a second service applies to blades in a second group, wherein the first and second services differ, and wherein the measurement values are taken from at least;

usage of processor resources, processing times, usage of memory, remaining capacity of data storage, and communication parameters of a blade interface; and

test the service in parallel operation on the earlier detected blade and the new blade;

cyclically repeating, in a first cycle, the copying, installing, and parallel testing for the plurality of blades in the first group, while keeping the number of blades in the first group that are re-installing the service smaller than the number of blades that are not re-installing the service, wherein a first specialized configuration is used for the service of the first group;

cyclically repeating, in a second cycle, copying, installing, and parallel testing for the plurality of blades in the second group while keeping the number of blades in the second group that are installing the service smaller than the number of blades in the second group that are not installing the service, wherein a second specialized configuration is used for the service of the second group; and

cyclically repeating, in a super cycle across the first and second groups, copying, installing, and parallel testing of a third service across the first and second groups, where the number of blades installing the third service is kept less than the number of blades not installing the third service,

wherein the copying of the second service is monitored by a group of blades other than the second group, and the copying of the first service is monitored by a group of blades other than the first group.

23. A tangible computer-readable storage medium containing instructions for execution by a processor for the practice of a method for managing a computer system, the instructions, when executed by the processor, causing the processor to:

assign a first service to a first group of computers and a second service to a second group of computers, wherein the first and second groups differ;

monitor system performance;

upon reaching a predefined threshold of a measurement value, shift a service that runs on a first computer of the group to run on a second computer in the group, wherein the shifting includes copying the first service from a memory of the first computer to the second computer, and wherein the measurement values are taken from at least;

usage of processor resources, processing times, usage of memory, remaining capacity of data storage, and communication parameters of a blade interface;

re-install a first service to the first computer;

cyclically repeat, in a first cycle, the shifting, installing, and parallel testing for the plurality of computers in the first group while keeping the number of computers in the first group that are re-installing the first service smaller than the number of computers that are not re-installing the first service, wherein a first specialized configuration is used for the first service of the first group;

cyclically repeat, in a second cycle, shifting, installing, and parallel testing for the plurality of computers in the second group while keeping the number of computers in the second group that are re-installing the second service smaller than the number of computers in the second group that are not re-installing the second service, wherein a second specialized configuration is used for the service of the second group; and

cyclically repeat, in a super cycle across the first and second groups, a shifting, installing, and parallel testing of a third service across the first and second groups, where the number of computers installing the third service is kept less than the number of computers not installing the third service,

wherein the shifting of the second service is monitored by a group of computers other than the second group, and the shifting of the first service is monitored by a group of computers other than the first group.