公开(公告)号：GB2520712A

公开(公告)日：2015-06-03

申请号：GB201321009

申请日：2013-11-28

Applicant: IBM

Inventor： SCHWEFLINGHAUS JOCHEN ,  LANG JAKOB ,  MENCIAS ANGEL NUNEZ ,  OLMOS VICTOR RAFAEL ESCOBAR ,  SCHIRMER ALBERT

IPC: G06F11/07

Abstract: During memory setup, computer memory is divided into primary 8 and secondary 9 sections, the primary section comprising first and second areas. Before a crash, first area 81 contains an Operating System (OS) and second area 84 contains an emergency OS and a dump program. The emergency OS and dump program are transferred from the second to the first area after the crash. The OS is operable before the crash and the emergency OS and dump program are operable after the crash. The secondary section is only indirectly controlled by the OS (widely out of its control). The first area contains a first SMSMP (Secondary Memory Section Management Program) which transfers first file system data 82 from the primary section to the secondary section before a crash. The second area contains a second SMSMP which directs read access of the dump program to the secondary section after a crash, in order to select second file system data 83 to be dumped using error data for the selection, e.g. an error code. The second SMSMP directs transfer of the emergency OS, dump program and second SMSMP code from the second to the first area.

公开(公告)号：GB2508344A

公开(公告)日：2014-06-04

申请号：GB201221368

申请日：2012-11-28

Applicant: IBM

Inventor： SCHWEFLINGHAUS JOCHEN ,  MENCIAS ANGEL NUNEZ ,  SCHIRMER ALBERT ,  LANG JAKOB ,  HOLZHEU MICHAEL ,  BUENDGEN REINHARD ,  SCHWIDEFSKY MARTIN

IPC: G06F11/07 ,  G06F11/14

Abstract: A method for creating an operating system dump comprises dividing a main memory of a computer system into at least three contiguous memory areas 102, comprising a primary memory area, a secondary memory area and a data memory area. Also comprises booting a first instance of an OS (operating system) into the main memory 104, loading a second instance of the OS into the secondary memory 106 area using the active first instance of the OS, stopping execution of the first active instance of the OS 108 if a critical execution error occurs and re-starting the computer using the loaded, second instance of the OS 110 which becomes the active instance of the OS. Method further comprises creating a dump of the primary memory area 112, and loading a third instance of the OS into the primary memory area 114. Thus the secondary instance continues regular operation of the primary instance, while dumping the crashed primary instance so it may be possible to minimize downtime after a crash.

公开(公告)号：GB2506195A

公开(公告)日：2014-03-26

申请号：GB201217077

申请日：2012-09-25

Applicant: IBM

Inventor： LANG JAKOB ,  MENCIAS ANGEL NUNEZ ,  SCHWEFLINGHAUS JOCHEN ,  SCHIRMER ALBERT

IPC: G06F9/455 ,  G06F9/50 ,  G06F21/53

Abstract: Method for managing virtual resources 20 on virtual machines (VMs) 10 by controlling execution of resources 20 by VM resource management layer 14 (VM instance) of trusted part 16 of a computer system, management layer 14 being unique in the computer system. VM instance 14 can be an interface between trusted part and VMs. Dynamic adding and releasing of VM resources by VM instance is defined. VM instance is on hypervisor, firmware layer, hardware layer and can be a license management system realized as billing system for using virtual resources. Billing information is securely transmitted to resource providers. Virtual resources 20 are software instances executed on a logical partition of a VM 10 and can be associated to physical resources for exclusive use. They can also be logical partitions, data processing units, main memories, network interfaces, resources of VMs. Accounting data for virtual resources are managed by VM instance 10.

公开(公告)号：GB2503439A

公开(公告)日：2014-01-01

申请号：GB201211276

申请日：2012-06-26

Applicant: IBM

Inventor： MENCIAS ANGEL NUNEZ ,  SCHIRMER ALBERT ,  SCHWEFLINGHAUS JOCHEN ,  LANG JAKOB

IPC: G06F11/14

Abstract: Disclosed is a method of restarting a software system 10 on a computer platform 12 after an unintentional software system shutdown. The method comprises generating a core dump file 14 on the computer platform when the software system shuts down unintentionally. Then restarting the software system, using a library 16 to copy software system related data 18 from the core dump file to the software system and continuing execution of the software system. The method may include the step of continuing execution of the software at a request of a request queue. The method may include checking the consistency of the software related data in the request and monitoring the software system for unintentional shutdown at the same request.