43 results about "Non-RAID drive architectures" patented technology

A JBOD subsystem for providing storage to a host entity. The JBOD subsystem contains at least one external JBOD emulation controller coupled to the host entity for emulating IO operations in response to the IO requests and a set of at least one physical storage device coupled to the JBOD emulation controller for providing storage to the host entity through the JBOD emulation controller. The JBOD emulation controller defines at least one logical media unit (LMU) consisting of sections of the set of physical storage device (PSD) and communicates with the PSD through a point-to-point serial-signal interconnect. As an example, the point-to-point serial-signal interconnect can be a Serial ATA IO device interconnect.

The invention discloses a hard disk sequence marshalling method based on a real-time JBOD monitoring mode. An order and a disk label of a hard disk are recorded through a server under the Linux operation system during process of each reboot (or startup and shutdown). A standard hard disk SN list is generated, so that obtained post processed data is compared with a standard hard disk SN list record, and the comparing result is saved. An alert is conducted when abnormality occurs during the data comparison. According to the method, besides the disk label sorting function of the hard disk can be monitored under a server system, a stable power supplying effect for a verification system can be achieved during the testing. The testing method has the advantages that the operation is simple, the utility range is wide, the portability is good, the practicality is high, and the stability of the server is effectively guaranteed. The testing method is an effective method for testing the quality of a product.

The invention discloses a design scheme of SAS Switch / JBOD topology. According to the scheme, a SAS Switch board and a control hard disk constitutes a storage structure which particularly comprises 1, a Rack computing server, 2, a SAS Switch node and 3, a JBOD node. According to the design scheme of the SAS Switch / JBOD topology, arrangement and pooling for computing source and storage resource are conducted through optimization of the whole structure, both solving expensive price of storage and dependence of storage server on server, so a novel structure takes shape, which meets application of client for distributed storage, mainly Hadoop structure, saves cost and is relatively high in flexibility, thus the amount of equipment can be flexibly adjusted according to demand of integral application.

The invention provides an SAN storage system and a data read-write method thereof. The SAN storage system comprises an application host, storage nodes and JBOD, each storage node comprises a control module and a caching module, each caching module comprises at least two caching blocks, each JBOD comprises at least two disk blocks, and the capacity of each caching block is the same as that of each disk block. The method comprises the steps that the control modules receive an SCSI command of the application host and control all the caching blocks to read and write data in all the disk blocks through a current caching strategy according to the SCSI command, the access frequencies of all the storage nodes are counted in real time, the data in the caching blocks of overloaded storage nodes is cached in the caching blocks of idle storage nodes, data read-write is conducted, a caching strategy matched with the access frequencies of all the storage nodes is selected every other preset time, and the selected caching strategy serves as the current caching strategy. The SAN storage system and the data read-write method thereof can be well applied to various application scenarios, the applicability is higher, cached resources can be effectively allocated, and the use rate of the cached resources can be increased.

The invention provides a JBOD (Just Bunch of Disks), which comprises a case, a controller and a disk, wherein the controller and the disk are arranged in the case; and the disk is connected with the controller. The JBOD can at least realize high density.

The invention discloses an N-way directly connected any to any controller architecture and a storage system architecture. Array controllers are configured with at least 2M ports, where M is an integers greater than two that corresponds to the number of SAS domains or JBOD units in the system. Serial attached SCSI (SAS) domains or JBOD units are configured with at least 2N ports, where N corresponds to the number of array controllers in the system. At least two of each of the 2N ports are directly connected to a corresponding at least two of each of the 2M ports of each of the N array controllers, thereby establishing direct redundant connectivity between each of the N array controllers and each of the M SAS domains or JBOD units.

A redundant storage system which can automatically recover RAID data by crossing different JBODs includes: at least one server, at least one SAS switch, and at least two storage devices; each of the at least one server includes an interface card, and each of the at least one server is connected to the at least one SAS switch through a SAS port of the interface card; each of the at least two storage devices is connected to the at least one SAS switch through a SAS interface; each of the at least two storage devices includes at least one physical disk; physical disks respectively included in different storage devices constitute a RAID group.

The invention discloses a testing method of the stability of a storage system with an expansion cabinet, relating to the field of tests of storage systems. The method comprises the following steps: a host computer BIOS is set to be electrified to be started and the host computer is set to be in shutdown at a fixed time; two fatigue machines are arranged and the fatigue machines are periodically electrified automatically; the first fatigue machine supplies power to the expansion cabinet JBOD and the second fatigue machine, and the second fatigue machine supplies power to the host computer; the host computer which is started is in shutdown at the fixed time, the first fatigue machine detects a host computer power failure signal, the first fatigue machine fails in power, the expansion cabinet JBOD fails in power and the second fatigue machine fails in power; after the first fatigue machine is electrified, the expansion cabinet JBOD is electrified, the second fatigue machine is electrified to supply power to the host computer, and after the host computer BIOS detects electrification, the host computer is started; and the storage system with the expansion cabinet is tested within a predetermined time in every electrification-failure period of the host computer, and a testing result is stored.

The invention relates to a testing method for Rack cabinet JBOD storage in the production environment. The testing method is characterized by including the specific steps that 1, a JBOD control server is filled with a system for server building; 2, basic function check is carried out, and the step 3 is carried out when basic function check is finished and no problem exists; 3, a restart stability test is carried out, and the step 4 is carried out when restart is finished and no problem exists after check; 4, JBOD storage equipment is subjected to pressure testing, and the step 5 is carried out when no problem exists in pressure testing; 5, basic function check is carried out, and the test is finished when no problem exists after check. Based on an open source Linux centos 7 system Shell language, a common testing tool is used for finishing production testing on Rack cabinet JBOD storage, unstable products or parts are discriminated, all quotative test items can be integrated into an integrated testing script, automatic start testing can be achieved, and efficiency is improved.

The invention discloses a JBOD device with a BMC module. The JBOD device mainly comprises a hard disk expansion control unit, a switching unit and a BMC module. The hard disk expansion control unit is connected with a number of hard disks. The switching unit is connected with the hard disk expansion control unit and a port to turn on a first transmission channel, or is connected with the hard disk expansion control unit and the BMC module to turn on a second transmission channel. The BMC module is connected with an Ethernet network through a network port. In the initial case, the hard disk expansion control unit receives an external wired transmission instruction through the first transmission channel. When the BMC module receives the external instruction transmitted through the Ethernet network, the switching unit is controlled to switch on the second transmission channel. The BMC module can interact with the hard disk expansion control unit through the instruction received through the Ethernet network.

Embodiments of the invention include a plurality of flexible electrical conductors configured as a cable wherein a plurality of signal pairs connect printed circuit boards in an array of data storage devices or just a bunch of disks (JBOD) enclosure. By controlling various specific dimensions relating to each signal pair of electrical conductors in a flexible cable, the performance of a JBOD box or a data storage server can be maximized. Furthermore, flexible cable designs can themselves replace bulkier circuit boards enabling greater air flow through the JBOD box or data storage server.

The invention discloses recognition system and method of just bundle of disks with a single-wire cascading topological structure, which are applied to a host with a plurality of controllers. The method comprises the following steps of: acquiring just bundle of disks (JBOD) correspondingly connected with a current controller of the host and hardware configuration information of a hard disk on the JBOD; synchronizing the hardware configuration information to other controllers of the host; and executing a plurality of controllers of the host to recognize the same JBOD in the single-wire cascading topological structure according to the synchronized hardware configuration information. The invention can realize the same recognition of the single-wire cascading JBOD under multi-controller environment.

The invention discloses a JBOD testing method and system based on a universal server, and relates to the technical field of server testing. The method comprises the steps that the serial number of a JBOD is verified, and verified serial number information is refreshed to an expander of the JBOD; Refreshing related division information of the JBOD according to the configured JBOD division version;Refreshing hard disk division information of the JBOD according to the related information of the hard disk; Performing a pressure test on the JBOD hard disk, and if the test has no problem, promptingthat the test is successful; And if the test has a problem, prompting that the test fails. On the basis of an existing general server test, the reliability test of the JOB can be realized without changing a test framework of the general server, the test method is simple and flexible, and the integrity and the accuracy of the test process of the server are improved.

The invention provides a storage system and a data transmission method. The storage system comprises a host and at least two JBODs which are sequentially arranged; according to the forward arranging sequence of the JBODs, a first output port of any one current JBOD is connected with a first input port of the next JBOD, a first input port of the first JBOD is connected with a main control port of the host, and then a main control link is formed; according to the reverse arranging sequence of the JBODs, a second output port of any one current JBOD is connected with a second input port of the next JBOD, a second input port of the first JBOD is connected with a standby control port of the host, and then a standby control link is formed; the host transmits JBOD access information through the main control link and the standby control link jointly when at least one of the JBODs works abnormally. According to the scheme, the data transmission reliability can be improved.

The invention relates the field of computer fittings, and specifically provides a low-cost JBOD, comprising a cabinet, a controller, a SAS RIAD card, a built-in SAS RIAD card external connection device and a hard disc, the hard disc connects with the controller by the SAS RAID card, the SAS RAID card and the SAS RAID card external connection device are connected. Compared with prior art, the low-cost JBOD has the characteristics of simple structure, and convenient usage, etc.

The embodiment of the invention discloses a method, a device and a system for sending a data packet, and exchanging equipment. The method comprises the following steps: the exchanging equipment receives an I / O data packet sent by upstream equipment of fault equipment when any one of at least two data links has a fault; a sending interface corresponding to an interface receiving the I / O data packet is confirmed according to the corresponding relationship between the preset interface receiving the I / O data packet and the sending interface; the I / O data packet is issued to the equipment corresponding to the fault equipment in the other data link through the sending terminal, and the other data link is the other one in the two data links except the data link in which the fault equipment is arranged. The embodiment of the invention realizes that the integrality of the entire I / O flow is maintained and the reliability of a double-control system or a multi-control system can be improved when modules and exchanging equipment which realize the same function in all controller or all JBODs do not simultaneously have faults.

The invention discloses an implementation method of disk arrays based on mixed media, and relates to storage field. According to the method, front-end interfaces are mutually connected with external servers through an extension network card, a rear-end device comprises a low-speed disk system and is connected with an SAS (serial attached SCSI) interchanger by the aid of JBOD (just a bunch of disks), the rear-end device is further connected with an SAS HBA (host bus adapter) and identified by the arrays, re-accessed response is accelerated by copying hotspot data in a disk storage, data correctness affected by faults of solid state disks is avoided, RAID 10 (redundant array of independent disks) operation of the solid state disks can be established when hybrid storage arrays are actually used, and fault tolerance of the solid state disks is improved.

The invention discloses a JBOD cascade system. The system comprises an SAS channel card and an SAS extended card set connected with the SAS channel card; and by utilizing a Zoning technology, an upper connection port of each SAS extended card in the SAS extended card set and a connected hard disk are both a same zone, and a lower connection port is a zone different from the upper connection port and the connected hard disk. By utilizing the Zoning technology, the upper connection port of each SAS extended card in the SAS extended card set and the connected hard disk are both the same zone, and the lower connection port is the zone different from the upper connection port and the connected hard disk, so that the SAS extended cards only can receive verification information of the SAS channel card through the upper connection ports and return reply information; and the SAS channel card judges whether the reply information is received or not to judge whether the SAS extended cards have a wiring error or not, so that fault hidden dangers caused by the wiring error are avoided.

The invention discloses an embedded private scale out network and a storage array system. N array controllers are configured with at least two interconnect fabric ports, where N is an integer greater than two. A JBOD unit is configured with at least two interconnect fabric switches. The two interconnect fabric switches each have at least N interconnect fabric ports. Each of the least N of the interconnect fabric ports are directly connected to a corresponding one of the N array controllers, thereby establishing direct redundant connectivity between each of the N array controllers and each other of the N array controllers.

The invention discloses a JBOD storage device testing method. The method comprises the steps of sending a shutdown command to each JBOD storage device connected to a general server; Setting the power-on self-starting parameter of each JBOD storage device to be in an open state through an interface connected with each JBOD storage device; Sending a power-off command to the PDU for supplying power to each JBOD storage device; Sending a power supply command to the PDU; And after the JBOD storage devices are restarted, testing the JBOD storage devices. According to the technical scheme provided bythe invention, in the process of testing the JBOD storage device, the waste of human resources is avoided by realizing the power-on self-starting of the JBOD storage device, and the test efficiency is greatly improved. The invention further provides a JBOD storage device testing system, a server and a computer readable storage medium which have the above beneficial effects.

The invention discloses a 1U universal modular high-density storage JBOD device. The device comprises a mounting housing, a storage control assembly which is inserted into the mounting space on one side of the mounting shell in a sliding manner and is used for storing data and processing instructions; a service interface assembly which is inserted into the mounting space on the other side of the mounting shell in a sliding manner and is used for realizing data interaction between the storage control assembly and external equipment; and a power supply assembly which is detachably mounted in themounting shell and is used for supplying power to the storage control assembly and the service interface assembly. According to the device, data interaction between external equipment and the storagecontrol assembly is realized under the action of the service interface assembly; in this way, data storage, reading and writing can be conveniently achieved, the storage control assembly, the serviceinterface assembly and the power supply assembly can be conveniently disassembled and assembled in the installation shell, maintenance and replacement of parts are facilitated, modular design and universal design of the parts are achieved, and the storage performance and the expansion performance of equipment are improved.

A server control method and a chassis controller are proposed. The server control method includes: determining whether the chassis controller is connected to a blade server when a power of the chassis controller is on; when the chassis controller is connected to the blade serve, further determining whether the blade server operates normally; when the blade server operates normally, switching the chassis controller to a slave device of the blade server, and controlling a just a bunch of disk (JBOD) system in response to a control command of the blade server.

The invention discloses a JBOD storage system. The JBOD storage system comprises a first-level SAS Expander, the first-level SAS Expander is provided with an expansion port for inward expansion, the expansion port is connected with at least two second-level SAS Expanders in parallel, and each second-level SAS Expander is connected with a storage part. By means of the JBOD storage system, all storage media in inside space can equally share bandwidth, storage reliability is improved, and performance decay of part of the storage media is reduced.

The invention discloses a method for guaranteeing security of JBOD mass memory data, and relates to the field of memory and security. The method comprises the steps that JBOD information security is realized by constructing an RAID, and the RAID enables an operating system to store data into and read data from a hard disk; an RAID system can recover data automatically according to verification fields on other hard disks; a virtual device is arranged between the server operating system and a JBOD device, when certain hard disk has a fault, the data in the failed hard disk is recovered by the aid of the verification fields, and the security performance of the JBOD is optimized through the virtual device. JBOD information security is realized through construction of the software RAID, the hard disk storage efficiency can be increased greatly without data losing; further, the security performance of the JBOD is optimized through the virtual device, the storage performance can be improved, and the data security can be improved.

A system and method are described for increasing performance of operations implemented in a storage system utilizing fibre channel, such as by enabling full duplex opens when a JBOD / bridge is involved. In an aspect of the present invention, a method of performing a full duplex open in a fibre channel network having an initiator and a bridge may include initiating an open by the initiator with the bridge. The bridge has a primary physical address and an alias physical address. The alias physical address represents a target device communicatively coupled to the bridge, wherein the open is initiated utilizing the primary physical address. The initiator communicates with the target device.

The FC optical fiber card comprises a mainboard connector, a signal conversion circuit and a plurality of output connectors, wherein the mainboard connector is used for being connected with a mainboard; the signal conversion circuit is used for dividing a PCIE signal of the mainboard into multiple paths of FC signals; wherein the total number of the output optical ports of the plurality of outputconnectors is a positive integer greater than 4; the mainboard connector is connected with the signal conversion circuit; the signal conversion circuit is respectively connected with each output connector, and each output connector corresponds to one FC signal sent by the signal conversion circuit; the PCIE signal of the mainboard is divided into multiple paths of FC signals by using the signal conversion circuit, through cooperation with the plurality of output connectors, the total number of the output optical ports of the FC optical fiber card is a positive integer larger than 4, the numberof the output optical ports of the FC optical fiber card is increased, the capacity of cascading JBOD of the FC optical fiber card is improved, the expansibility is better, and meanwhile, the size ofthe FC optical fiber card can be ensured not to be increased.

The invention provides an overall hot backup method for multiple controllers. In the method, a high-end eight-control disc array is composed of eight controllers; each controller is provided with a front end card and a rear end card; the front end cards are 10-gigabit Ethernet cards, and are used for connecting with a front end host machine or various servers; the rear end cards are SAS (Serial Attached SCSI) cards, and are connected with a SAS switch; the SAS switch is connected with JBOD (Just Bundle Of Disks) discs; each controller is correspondingly connected with three JBOD discs and an overall hot backup JBOD disc; and the hot backup JBOD discs are shared for the overall storage system. According to the invention, the reliability of the storage system is improved, the data security is ensured, the utilization rate of stored resources is improved, and the enterprise cost is greatly reduced.

A method for auto recovery storage of JBOD array is disclosed. The method includes: determining whether a disk of the JBOD array is failed, deleting a storage resource according to a storage resource list for the failed disk is stored with an index area upon determining the failed disk, updating the index area corresponding to a recorded data area when the failed disk is stored with the storage resource of the recorded data area instead of the index area; transmitting control instructions indicative of adding a hot spare to the JBOD array to add the hot spare to the JBOD array, and adding the storage resource and activating the storage resource for the failed disk stored with the index area after the hot spare is added to the JBOD array.

The invention provides a storage system and a data transmission method. The storage system comprises a host and at least two JBODs which are sequentially arranged; according to the forward arranging sequence of the JBODs, a first output port of any one current JBOD is connected with a first input port of the next JBOD, a first input port of the first JBOD is connected with a main control port of the host, and then a main control link is formed; according to the reverse arranging sequence of the JBODs, a second output port of any one current JBOD is connected with a second input port of the next JBOD, a second input port of the first JBOD is connected with a standby control port of the host, and then a standby control link is formed; the host transmits JBOD access information through the main control link and the standby control link jointly when at least one of the JBODs works abnormally. According to the scheme, the data transmission reliability can be improved.