METHOD FOR OPERATING A NODE CLUSTER SYSTEM IN A NETWORK AND NODE CLUSTER SYSTEM

    公开(公告)号:CA2753747C

    公开(公告)日:2019-08-13

    申请号:CA2753747

    申请日:2011-09-30

    Applicant: IBM

    Abstract: The invention relates to a method for operating a node cluster system with a plurality of nodes in a network, wherein the cluster system (100) appears to be a single node with only one specific network address (IP, MAC) to its network environment. The method comprising the steps: providing a shared socket database for linking network connection port identifications of a common set of network connection port identifications to the individual nodes, assigning a master function to one of the nodes, sending incoming traffic to all nodes of the cluster system, wherein each node verifies its responsibility for this traffic individually, exclusive assignment of a network connection port to the responsible node for the duration of a connection of the corresponding application process by means of the corresponding network connection port identification and the link established by the shared socket database, and processing of the traffic by the responsible node or otherwise by the node having the master function. Further, the invention relates to a corresponding computer-readable medium, to a corresponding computer program product and to a corresponding node cluster system.

    Migration of virtual machines with reassignment of hardware memory segments

    公开(公告)号:GB2508231A

    公开(公告)日:2014-05-28

    申请号:GB201221265

    申请日:2012-11-27

    Applicant: IBM

    Abstract: A virtual machine (VM) is executed 200 on a source virtual machine monitor (VMM, e.g. hypervisor) 110 within a logical partition. The VM is allocated certain segments of hardware memory 201 assigned to the partition for exclusive data storage. When the VM is to be migrated to a destination VMM 120, the operation of the machine is suspended 202, and information relating to the configuration of the VM, including the allocated memory segments, is sent to the new VMM 203. The memory segments are then reassigned to the logical partition containing the destination VMM 206. The VM is then shut down 205 on the source VMM, with extended configuration information including descriptions of the assigned memory segments used to recreate the VM at the destination VMM, and the operation of the VM is then resumed on the new VMM 207. This allows the transfer of the data created by the virtual machine without requiring duplication of content and/or excessive processing time.

    Accessing privileged objects in a server by means of semi privileged instruction

    公开(公告)号:GB2507339A

    公开(公告)日:2014-04-30

    申请号:GB201219367

    申请日:2012-10-29

    Applicant: IBM

    Abstract: Method comprising: associating privileged objects (socket connection, disk-device, input/output bus access) 32.1-32.2 with application 20 comprising process resources 22.1-22.2 and corresponding semi-privileged instruction; filing the association in entity (associating table) 34 of operating system kernel 30; wherein central processing unit (CPU) 40 performs authorization check 42 if semi-privileged instruction attempts accessing privileged objects; wherein CPU executes 44 the semi-privileged instruction granting access to privileged objects if the kernel issued the semi-privileged instruction; or accesses kernel-entity if resources 22.1-22.2 issued the instruction to determine their authorization to access privileged objects; upon positive authorization, CPU executes the semi-privileged instruction granting access to privileged objects; upon failure, CPU denies execution of instruction and performs 46 corresponding authorization failure handling. File descriptor is used: by semi-privileged instruction to transfer data; by user space library to block, use select, poll, epoll; by kernel to verify resource authorization to use data transfer connection if forking occurs.

    Migration of a virtual machine between hypervisors

    公开(公告)号:GB2506684A

    公开(公告)日:2014-04-09

    申请号:GB201217986

    申请日:2012-10-08

    Applicant: IBM

    Abstract: The invention relates to a method for replacing a source hypervisor (130.1) by a target hypervisor (130.3) in a system (100) comprising a first level hypervisor (121) managing multiple logical partitions (101, 103, 105) to share a set of resources of a server (125). The source hypervisor (130.1) enables at least one source virtual machine (106.1) to share said resources that are assigned to a source logical partition of the multiple logical partitions where the source hypervisor (130.1) runs. The target hypervisor (130.3) enables at least one target virtual machine (106.5) to share said resources that are assigned to a target logical partition of the multiple logical partitions where the target hypervisor (130.3) runs, wherein memory segments (206.1) of a physical memory (123) of the server (125) are assigned to the at least one source virtual machine (106.1). A memory map is used to transfer the memory state across the virtual machine managers.

Patent Agency Ranking