摘要:
In a parallel processor system comprising a plurality of processor elements constituting a network, a source processor element wishing to broadcast data to a plurality of destination processor elements sends a broadcast request message containing the target data to a broadcast exchanger. The broadcast exchanger converts the received message into a broadcast message and sends it over the network to the destinations. A plurality of broadcast request messages, if transmitted parallelly to the broadcast exchanger, are serialized thereby so that only one broadcast message will be transmitted at a time over the network. This prevents deadlock from occurring between different broadcast messages. The routes for transmitting broadcast request messages and those for transmitting broadcast messages are arranged so as not to overlap with one another. This suppresses deadlock between any broadcast request message and broadcast message. The broadcast exchanger is replaced alternatively with one of the partial networks. These schemes all apply where long messages are transmitted through worm-hole routing.
摘要:
In a multi-tenancy information processing system, a necessary setting change item and a target NW device are specified at the time of changing a configuration of a virtual machine of a tenant. A management server 116 includes a device management table 352 for each tenant and each segment, in which an NW setting item set for a network (NW) device is stored for each of the NW devices corresponding to the segment of the tenant, in advance. The management server 116 receives a configuration change request including type information on a request representing addition, deletion, or movement of the virtual machine, an identifier of a target tenant, and an identifier of a target segment.
摘要:
In a multi-tenant information processing system, when initially constructing a tenant, design items depending on the tenant are specified for each device constituting the tenant, and design values of the design items are calculated without overlapping with those of other tenants. In a tenant design item management table, the design items set for the devices and information indicating whether the design items depend on the tenant, are stored for each device. In a design algorithm management table, calculation formulae for making the design values to be different for each tenant are stored corresponding to the design items. A tenant design unit refers to the tenant design item management table, specifies the design items depending on the tenant among those corresponding to the devices constituting the tenant, calculates the design values of the design items according to the calculation formula, when initially constructing the tenant.
摘要:
A method of monitoring a virtual computer system including a network device coupled to a client computer, sever computers each coupled to the network device, and a performance degradation source locator apparatus coupled to the network device and the server computer, obtaining, by the performance degradation source locator apparatus, performance information on physical computer resources and performance information on virtual computer resources of extracted virtual machines and server computers and the virtual machine and the server computer in which the latency has increased to locate a part in which the latency has increased.
摘要:
A failure management method for a computer including a processor, and a memory connected to the processor, and in which the processor containing a memory protection function, executes a first software program and a second software program monitoring the operation of the first software program, and the second software program retains error information including address information and access-related information; and the method implemented by the by the second software program includes a step for detecting the occurrence of errors in the memory; and a step for prohibiting access to the address of the memory where the error occurred, and monitoring the access state; and a step for executing the failure processing when accessing by the first software program of the address of the memory where the error occurred was detected.
摘要:
(1) A configuration management in a management server is carried out such that setting patterns regarding creation of plural types of virtual resources and setting patterns regarding connection between the plural types of virtual resources are extracted from setting information of network devices configuring a virtual network by using plural types of virtualization technologies, (2) an ID required for creating a virtual resource and a connection relation between virtual resources are extracted from the extracted setting pattern, (3) an element corresponding to a virtual resource or a connection between a pair of virtual resources is created by using information including the extracted ID, and (4) the elements created from setting information of individual devices are aggregated to thereby create logical topology information representing a topology of the virtual network.
摘要:
A computer system reduces an overhead by using an I/O switch high in versatility when realizing the sharing of an I/O between virtual computers. The system includes a CPU module #0 having plural CPU cores, an AS bridge 15 connected to the CPU cores, and a main memory that can be accessed from the CPU cores or the AS bridge 15, and AS switches SW0 and SW1 that connect the AS bridge 15 of the CPU module #0 and an I/O blade #5. The CPU module #0 has a hypervisor that divides the plural CPU cores and the main memory into plural logical partitions, and the AS bridge 15 adds virtual path information set in each of the logical partitions and path information of from the AS bridge 15 to the I/O blade #5 to the path information of the AS packet to switch over the AS packet between each of the logical partitions and the I/O blade #5 when relaying the AS packet transmitted or received between the logical partitions and the I/O blade #5.
摘要:
There are provided a heap memory management method that not only improves processing performance of an application program but also reduces a program development cost and a computer system using the same method. A live object copy process in minor collection of generational garbage collection is implemented through a calculate process of object live ratio, a compare process of object live ratio, and a live object copy process. The calculate process of object live ratio calculates a ratio of a size of a memory allocated to live objects of new objects to a size of a memory allocated to a new object during minor collection. The compare process of object live ratio compares a calculated object live ratio with a predetermined threshold. The live object copy process copies a live object directly to an old-generation heap instead of a new-generation heap when the object live ratio exceeds the threshold as a result of comparison. Further, an application programmer provides means for setting the threshold that decides a copy destination of the live object as a runtime parameter of a program and means for holding the set threshold in a memory.
摘要:
Provided is a fibre channel switch system to which a server and a storage system are connected. The fibre channel switch system includes: a host controller for controlling a fibre channel protocol, to which the server is connected; a management table for indicating a hardware address of the host controller; a switching unit for routing information of the fibre channel protocol; and a control unit for controlling the host controller and the switching unit. Accordingly, the server connected to fibre channel switch system can be downsized.
摘要:
In a step for bonding a color filter to a solid-state color-image sensor chip with an adhesive which is curable not only by light or more particularly ultraviolet-ray irradiation but also by heating, pre-curing or partial curing is effected by irradiating light rays after the color filter and the chip has been pressed against each other and correctly aligned with each other in a bonding device and then the chip with the partially-bonded color filter is removed from the bonding device and complete curing of the adhesive is accomplished by heating. According to one embodiment of the present invention the bonding step is carried out in an atmosphere containing the oxygen so that the adhesive which has been squeezed out from the space between the color filter and the chip may be prevented from being cured and subsequently removed in a simple manner. According to a further embodiment of the present invention, prior to the bonding step, the color filter and chip are covered with protective layers or films so that surface flaws may be avoided. Thus, the color filters can be bonded to the chips with a higher degree of accuracy and a higher degree of adhesive strength and the mass production of solid-state color-image sensors can be much facilitated.