CN101178634A - Knife blade server and storage realization method therefor - Google Patents
Knife blade server and storage realization method therefor Download PDFInfo
- Publication number
- CN101178634A CN101178634A CNA2006101143830A CN200610114383A CN101178634A CN 101178634 A CN101178634 A CN 101178634A CN A2006101143830 A CNA2006101143830 A CN A2006101143830A CN 200610114383 A CN200610114383 A CN 200610114383A CN 101178634 A CN101178634 A CN 101178634A
- Authority
- CN
- China
- Prior art keywords
- memory controller
- memory
- blade server
- bus
- disc blades
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims description 21
- 230000005540 biological transmission Effects 0.000 claims description 8
- 239000000835 fiber Substances 0.000 claims description 5
- 230000002093 peripheral effect Effects 0.000 claims description 4
- 230000005855 radiation Effects 0.000 abstract description 4
- 230000009286 beneficial effect Effects 0.000 abstract 1
- 230000006870 function Effects 0.000 description 9
- 238000013461 design Methods 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000005856 abnormality Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 239000012141 concentrate Substances 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- RGNPBRKPHBKNKX-UHFFFAOYSA-N hexaflumuron Chemical compound C1=C(Cl)C(OC(F)(F)C(F)F)=C(Cl)C=C1NC(=O)NC(=O)C1=C(F)C=CC=C1F RGNPBRKPHBKNKX-UHFFFAOYSA-N 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012797 qualification Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Images
Landscapes
- Multi Processors (AREA)
Abstract
The invention discloses a blade server, comprising a memory cell. The invention is characterized by a disk free blade and a memory controller further comprised, wherein, the memory controller is used for managing the memory cell and accomplishing the memory access operation of the memory cell; and the disk free blade is connected with the memory controller through a memory bus, so as to accomplish a memory access on the memory cell by the memory controller. The invention has the advantages that the blade server achieves the concentrated memory, thereby being beneficial to the thermal radiation of the disk free blade and being easy to achieve the thermal plug in and out of the hard disk; besides, when the blade serve accomplishes the RDA on the hard disk data, the invention improves the utilization rate of the hard disk and the memory capacity of the blade server.
Description
Technical field
The present invention relates to server and field of storage, relate in particular to a kind of blade server and a kind of storage implementation method that is applied to this blade server.
Background technology
Blade server has characteristics such as low-power consumption, the space is little, the unit price is low, and these characteristics have satisfied the demand of intensive computing environment, thereby have obtained increasingly extensive application.Blade server adopts the rack cabinet of calibrated altitude usually, and comprising backboard, but the server unit of a plurality of cassettes of plug-in mounting on the backboard is called blade, to realize high available and high density.Each piece blade all comprises mainboard, CPU (CentralProcess Unit, central processing unit) etc., and operating system that can independent operating oneself is equivalent to an independently server on the function.Certainly, the keeper also can be the server cluster operation with a plurality of blade configuration.
In the prior art, the mode that each blade of blade server adopts main board system and hard disk to unify, and at first with the hard disk installed on this blade as its storage space.
One skilled in the art will appreciate that the hard disk volume is bigger, the heat radiation of understanding blade impacts, and the hot plug of hard disk realizes understanding the degree of freedom when reducing insert design.Because the area of blade is subjected to the restriction of frame, and two hard disks often can only be installed on the blade.Realize the redundancy of hard disc data, usually two hard disks are configured to RAID (Redundant Array of Inexpensive Disk, redundant arrays of inexpensive disks) 1, promptly two hard disks are carried out mirror image, spendable like this storage space is equivalent to the capacity of a hard disk, and the utilization factor of hard disk has only 50%; Simultaneously, it is limited to be configured to the performance that the hard disk behind the RAID 1 can reach, and the performance with a hard disk is suitable at the most when write operation, can not surpass when read operation simultaneously from the speed of the reading of data of two hard disks.
Summary of the invention
The present invention will solve is because of blade combines with hard disk that the hard disk utilization factor that causes is low, poor-performing and the bad problem of heat radiation in the prior art.
Blade server of the present invention comprises storage unit, and does not have disc blades and memory controller, wherein:
Memory controller is used for the managed storage unit, and finishes the storage access operations to storage unit;
No disc blades is connected with memory bus with memory controller, by memory controller storage unit is carried out memory access.
Preferably, described blade server also comprises backboard, connects no disc blades and memory controller respectively, is used for realizing not having disc blades and is connected with memory bus between memory controller; The connecting interface of this backboard and no disc blades and memory controller comprises the signal wire of memory bus.
Preferably, described blade server also comprises administrative unit, connects memory controller by management bus, is used for memory controller is configured and monitors.
Alternatively, described memory bus is parallel memory bus.
Alternatively, described memory bus is a universal serial bus;
Described blade server also comprises crosspoint, is connected between no disc blades and the memory controller by memory bus, is used for transmitting between no disc blades and memory controller memory access information.
Preferably, described blade server also comprises administrative unit, connects crosspoint and memory controller by management bus, is used for crosspoint and memory controller are configured and monitor.
Preferably, described blade server also comprises backboard, connects no disc blades and crosspoint respectively, is used for realizing not having disc blades and is connected with memory bus between crosspoint; The connecting interface of this backboard and no disc blades and crosspoint comprises the signal wire of memory bus.
Alternatively, described blade server also comprises the backboard that is integrated with crosspoint, and the connecting interface of this backboard and no disc blades and memory controller comprises the signal wire of memory bus.
Alternatively, described memory bus comprises serial attached small computer system interface SAS, serial AT computing machine optional equipment SATA, internet attached small computer system interface iSCSI, fiber optic protocols interface FCP, quick peripheral assembly interconnecting PCI-E or Infiniband bus.
Preferably, described memory controller comprises expansion just a bunch of disks JBOD hard disk cabinet interface.
Preferably, described storage unit is a hard disk.
The invention provides a kind of storage implementation method that is applied to described blade server, may further comprise the steps:
Memory controller is divided at least one logical block with the storage unit of its management;
Memory controller is set up the logic connecting relation between logical block and the no disc blades;
No disc blades is carried out memory access by memory controller to the logical block that its logic connects.
Preferably, described method also comprises before memory controller is set up logic connecting relation between logical block and the no disc blades: no disc blades obtains device identification ID according to bus protocol, sets up logic connecting relation according to this and finishes memory access for memory controller.
Preferably, to comprising the blade server of administrative unit, described method also comprises: administrative unit manages it according to the duty of memory controller transmission.
Among the present invention, main board system on the blade is separated with hard disk, make no disc blades, make blade server realize centralized stores by can storage unit being conducted interviews shared memory controller, help not having the heat radiation of disc blades, be easy to realize the hot plug design of hard disk; Simultaneously, when carrying out the hard disc data redundant storage, improved blade server the utilization factor of hard disk and the memory property of blade server.
Further, the present invention is configured and monitors memory controller and/or crosspoint by making administrative unit, has realized the centralized management of storing in the blade server.
Description of drawings
Fig. 1 is the structural representation of blade server embodiment one of the present invention;
Fig. 2 is the structural representation when memory bus is by the backboard cabling among the blade server embodiment one of the present invention;
Fig. 3 is the process flow diagram of blade server embodiment one storage implementation method of the present invention;
Fig. 4 is the structural representation of blade server embodiment two of the present invention;
Fig. 5 is the structure intention when memory bus is by the backboard cabling among the blade server embodiment two of the present invention;
Fig. 6 is the process flow diagram of blade server embodiment two storage implementation methods of the present invention.
Embodiment
Basic design of the present invention is that the hard disk in the blade is separated with remainder, and concentrates in blade server and realize hard-disc storage, these storage spaces is divided the back use for each no disc blades.No disc blades alleged among the present invention can be the blade of the various blade servers behind the removal hard disk for not comprising the blade of hard disk.Describe embodiments of the invention in detail below in conjunction with accompanying drawing.
In embodiments of the invention one, adopt the blade server of serial memory bus can have structure shown in Figure 1.No disc blades 111,112 and 113 is connected to crosspoint 160 by the serial memory bus, memory controller 130 connects crosspoint 160 and storage unit 141,142 and 143 respectively, and administrative unit 150 is connected with memory controller 130 with crosspoint 160 by management bus.Storage unit in the present embodiment comprises hard disk, can also comprise other memory devices such as magnetic tape station.
Among the present invention, the storage unit in the blade server is connected to memory controller 130, carries out centralized configuration by memory controller 130, the storage space after the configuration is offered each no disc blades respectively use.For example, the storage space of whole blade server can be divided into different subregions, each no disc blades is used wherein at least one subregion.Memory controller 130 also is responsible for finishing the memory access to storage unit, and when no disc blades conducted interviews to the storage space of distributing to its use, the accessing operation instruction arrived memory controller 130; Memory controller 130 will be carried out after will being converted to accessing operation to storage unit to the accessing operation of storage space, and execution result is returned.
When the same time in the blade server was only used a no disc blades, no disc blades can directly be connected with memory controller 130.When using in the blade server simultaneously,, need between no disc blades and memory controller 130, increase crosspoint 160 based on the characteristic of serial memory bus above a no disc blades.Crosspoint 160 is connected by the serial memory bus respectively with memory controller with no disc blades, coordinate of the visit of a plurality of no disc blades to storage unit that memory controller 130 connects, the memory access information of each no disc blades is forwarded to memory controller 130, and the memory access execution result information of memory controller 130 is forwarded to the no disc blades of initiating this memory access.
In the blade server, blade is plugged on the backboard by slot interface usually, by backboard realize with blade server in being connected of miscellaneous part.Among the present invention, in order to realize not having the integral joint of disc blades, can in the connecting interface of no disc blades and backboard 120, increase the signal wire of memory bus, in the connecting interface of crosspoint 160 or memory controller 130 and backboard 120, also increase simultaneously the signal wire of memory bus, with no disc blades and crosspoint 160 or do not have disc blades and memory controller 130 between memory bus be connected realization on the backboard 120.Like this, no disc blades, crosspoint 160 or memory controller 130 can insert backboard 120 by a whole interface, have simplified the connection design between the integrated circuit board.When adopting this mode, do not have in the blade server between disc blades, backboard 120 and the memory controller 130 annexation as shown in Figure 2.
Need to prove that crosspoint 160 can be integrated on the backboard 120 and realize.At this moment, no disc blades and memory controller 130 insert backboard 120 by the connecting interface that comprises the memory bus signal wire respectively, are connected with crosspoint 160.
Please again referring to Fig. 3, when needs manage and are provided with memory controller and/or crosspoint 160, can increase administrative unit 150.Administrative unit 150 is connected to memory bus controller 130 and crosspoint 160 by management bus, by the storage unit of the configuration of memory controller 130 and crosspoint 160 and monitoring being come be connected on the managing blade server.In the prior art, usually also comprise management bus and the administration module that miscellaneous part is managed in the blade server, in the function of this administration module, increase management function to memory controller 130 and crosspoint 160 and can realize administrative unit 150 among the present invention.Management function to memory controller 130 and crosspoint 160 can adopt the storage management technique that is complementary with employing serial memory bus in the prior art.Can adopt universal serial bus as management bus among the present invention, as I2C (Inter-Integrated Circuit, intraconnection integrated circuit) bus or RS485 (No. 485 proposed standards) bus.
Consider the memory capacity of blade server, can select the memory controller 150 with expansion interface for use, blade server can pass through the external expansion disk chassis of this expansion interface, and bigger memory capacity is provided during memory space inadequate in blade server.The present invention recommends to adopt expansion JBOD (Just a BunchOf Disks, just a bunch of disks) disk chassis.
Serial memory bus in the present embodiment can be iSCSI (Internet Small Computer SystemInterface, internet attached small computer system interface), PCI-E (Peripheral Component InterconnectExpress, quick peripheral assembly interconnecting), SATA (Serial Advanced Technology Attachment, serial AT computing machine optional equipment), SAS (Serial Attached SCSI, serial attached small computer system interface), FCP (Fiber Distributed Digital Interface, fiber optic protocols interface), Infiniband bus and other.Wherein, the Infiniband bus claims that also the Infiniband framework is a kind of Convertion cable technology of supporting how concurrent link.In this technology, every kind of link can reach the travelling speed of 2.5Gbps (lucky bits per second).This framework speed in a link is 500MB/ second (megabyte per second), and speed is 2GB/ second (giga bits per second) in the time of four links, and speed can reach 6GB/ second in the time of 12 links.The InfiniBand technology is not used for general networking and connects, and its main purpose of design is at the connectivity problem of server end.Therefore, the InfiniBand technology will be applied to communicating by letter between server and server, server and memory device and server and the network.
When adopting iSCSI, crosspoint 160 can be an Ethernet switch, and transmission iSCSI signal between the no disc blades, and the interface between the memory controller 130 is an ethernet network interface; Memory controller 130 can connect SATA, SCSI (Small Computer Systems Interface, small computer system interface) or FC (Fibre Channel, optical-fibre channel) hard disk, the expansion interface of memory controller 130 can be SCSI, RJ45 (a kind of twisted-pair feeder standard interface), LC (a kind of sonet standard interface), HSSDC (High-Speed Serial Data Connector, high speed data interface).
When adopting the PCI-E bus, crosspoint 160 can be the PCI-E switch, and transmission PCI-E signal between the no disc blades, and the interface between the memory controller 130 is the PCI-E interface; Memory controller 130 can connect SATA, SCSI or FC hard disk, and the expansion interface of memory controller 130 can be SCSI, RJ45, LC, HSSDC interface.
When adopting the SATA bus, crosspoint 160 can be the SATA switch, and transmission SATA signal between the no disc blades, and the interface between the memory controller 130 is the SATA interface; Memory controller 130 can connect SATA, SCSI or FC hard disk, and the expansion interface of memory controller 130 can be SCSI, RJ45, LC, HSSDC interface.
When adopting the SAS bus, crosspoint 160 can be the SAS switch, and transmission SAS signal between the no disc blades, and the interface between the memory controller 130 is the SAS interface; Memory controller 130 can connect SATA, SCSI or FC hard disk, and the expansion interface of memory controller 130 can be SCSI, RJ45, LC, HSSDC interface.
When adopting FCP, crosspoint 160 can be the FCP switch, and transmission FCP signal between the no disc blades, and the interface between the memory controller 130 is the FCP interface; Memory controller 130 can connect SATA, SCSI or FC hard disk, and the expansion interface of memory controller 130 can be SCSI, RJ45, LC, HSSDC interface.
When adopting the Infiniband bus, crosspoint 160 can be the Infiniband switch, and transmission Infiniband signal between the no disc blades, and the interface between the memory controller 130 is the Infiniband interface; Memory controller 130 can connect SATA, SCSI or FC hard disk, and the expansion interface of memory controller 130 can be SCSI, RJ45, LC, HSSDC interface.
To having the blade server of structure shown in Figure 2, it preferably stores implementation method can adopt flow process shown in Figure 3.
Step S301: insert no disc blades.
Step S302: no disc blades is surveyed the connection status with the blade server backboard, if connection status is normal, and execution in step S303; If note abnormalities, then realize sending warning by hardware by itself.
Step S303: no disc blades joins in the topological structure of universal serial bus storage with the form of an equipment.
Step S304: after no disc blades adds universal serial bus memory topology structure,, send application or broadcasting, to obtain a device id unique in this topological structure (sign) according to the agreement that the serial memory bus is adopted.
Step S305: no disc blades is logined in crosspoint or is registered, and preserve its device id after obtaining the unique ID of equipment, and other device attribute information, as transfer rate etc.The device id that each no disc blades is stored in crosspoint is transparent to memory controller, and promptly memory controller can be learnt each no disc blades and corresponding equipment ID thereof.
Step S306: the memory cell arrangements that memory controller connects is a logical block.For example, according to the difference of applied environment, memory controller can be set to different RAID ranks, and a plurality of storage unit are bound into the RAID group; Memory controller is further divided the RAID group, is divided into the logical block of one or more support os startings; The quantity of logical block is no less than the quantity of no disc blades.
Step S307: memory controller is in logical block and do not have and set up logic connecting relation between the disc blades.Like this, under various operating systems, logical block and local hard drive that no disc blades is seen do not have difference.
Step S308: no disc blades is carried out memory access by memory controller to the logical block that its logic connects.
Step S309: crosspoint, the common management agreement of memory controller support, self working state is delivered to management bus, arrive administrative unit through management bus.
Step S310: administrative unit is distinguished the management data that receives from management bus and is belonged to crosspoint or memory controller by the type of management bus and crosspoint and memory controller; And crosspoint and memory controller are managed according to the management data that receives.
In embodiments of the invention two, blade server adopts parallel memory bus, and its structural representation as shown in Figure 4.No disc blades 411,412 and 413 is connected to memory controller 430 by parallel memory bus, and memory controller 430 connects storage unit 441,442 and 443, and administrative unit 450 is connected with memory controller 430 by management bus.
Storage unit in the blade server is connected to memory controller 430, carries out centralized configuration by memory controller 430, the storage space after the configuration is offered each no disc blades respectively use.For example, the storage space of whole blade server can be divided into different volumes and distribute to each no disc blades.Memory controller 430 also is responsible for finishing the memory access to storage unit, when no disc blades conducts interviews to the storage space of distributing to its use, memory controller 430 receives the memory reference order of no disc blades, carry out after will being converted to accessing operation to the accessing operation of storage space, and execution result is returned storage unit.
In the blade server, blade is plugged on the slot interface of backboard usually, realize with blade server in being connected of miscellaneous part.Among the present invention, in order to realize not having the integral joint of disc blades, can in the connecting interface of no disc blades and backboard 420, increase the signal wire of parallel memory bus, in the connecting interface of memory controller 430 and backboard 420, also increase the signal wire of parallel memory bus simultaneously, the memory bus between no disc blades and the memory controller 430 is connected on the backboard 420 realizes.Like this, no disc blades, memory controller 430 can insert backboard 420 by a whole interface.When adopting this mode, do not have in the blade server between disc blades, backboard 420 and the memory controller 430 annexation as shown in Figure 5.
Please, if desired memory controller is managed and is provided with, can increase administrative unit 450 again referring to Fig. 4.Administrative unit 450 is connected to memory bus controller 430 by management bus, manages the storage unit that it is connected by the configuration to memory controller 430 with monitoring.In the prior art, also comprise management bus and the administration module that miscellaneous part is managed usually in the blade server, in the function of this administration module, increase management function to memory controller 430 and can realize administrative unit 450 among the present invention.Management function to memory controller 430 can adopt storage management technique of the prior art to realize.Can adopt universal serial bus as management bus among the present invention, as I2C bus or RS485 bus.
In order the storage expanded function of blade server to be provided, can to select for use the memory controller 450 with expansion interface, blade server is by the external expansion disk chassis of this expansion interface, and bigger memory capacity is provided during memory space inadequate in blade server.The present invention recommends to adopt expansion JBOD disk chassis.
Parallel memory bus in the present embodiment recommends to adopt SCSI bus, and this moment, memory controller 430 can connect storage unit by SAS or scsi interface, and the expansion interface of memory controller 430 also can adopt SAS interface or scsi interface.Storage unit in the present embodiment comprises hard disk, can also comprise other memory devices such as magnetic tape station.
To having the blade server of structure shown in Figure 5, it preferably stores implementation method can adopt flow process shown in Figure 6.
Step S601: insert no disc blades.
Step S602: no disc blades is surveyed the connection status with the blade server backboard, if connection status is normal, and execution in step S603; If note abnormalities, then realize sending warning by the hardware of itself.
Step S603: no disc blades joins in the topological structure of parallel bus storage with the form of an equipment.
Step S604: after no disc blades added parallel bus memory topology structure, according to the agreement that parallel memory bus is adopted, the application of giving an order was to obtain a device id unique in this topological structure.The device id of each no disc blades is transparent to memory controller, and promptly memory controller can be learnt each no disc blades and corresponding equipment ID thereof.
Step S605: the memory cell arrangements that memory controller connects is a logical block.For example, according to the difference of applied environment, memory controller can be set to different RAID ranks, and a plurality of storage unit are bound into the RAID group; Memory controller is further divided the RAID group, is divided into the logical block of one or more support os startings; The quantity of logical block is no less than the quantity of no disc blades.
Step S606: memory controller is in logical block and do not have and set up logic connecting relation between the disc blades.Like this, under various operating systems, logical block and local hard drive that no disc blades is seen do not have difference.
Step S607: no disc blades is carried out memory access by memory controller to the logical block that its logic connects.
Step S608: memory controller is delivered to management bus according to management agreement with self working state, arrives administrative unit through management bus.
Step S609: administrative unit receives the management data of memory controller from management bus, and according to the management data that receives memory controller is managed.
Because the present invention has realized the centralized management of storage, in a blade server, only need a spot of memory controller, under the condition that does not increase the blade server cost, memory controller be can select for use, thereby the memory access performance of blade server and the data security of hard disk improved with more powerful memory management functions.
Even do not consider the performance of memory controller, after using the present invention, the hard disk quantity of centralized management all surpasses two usually, can be configured as RAID 5, N hard disk is configured to RAID 5 its utilization factors of back and is (N-1)/N, 50% utilization factor when far being RAID 1 above the blade configuration that has hard disk; Simultaneously, after being configured to RAID 5, be that read operation or write operation all are to carry out simultaneously at the polylith hard disk, promoted the memory property of blade server.
Same realization owing to centralized stores, blade server can be realized the redundancy of hard disc data with lower cost, helps ensureing the safety of data.In addition, no disc blades has better radiating effect.
Above-described embodiment of the present invention does not constitute the qualification to protection domain of the present invention.Any modification of being done within the spirit and principles in the present invention, be equal to and replace and improvement etc., all should be included within the claim protection domain of the present invention.
Claims (14)
1. a blade server comprises storage unit, it is characterized in that, also comprises no disc blades and memory controller, wherein:
Memory controller is used for the managed storage unit, and finishes the storage access operations to storage unit;
No disc blades is connected with memory bus with memory controller, by memory controller storage unit is carried out memory access.
2. blade server as claimed in claim 1 is characterized in that described blade server also comprises backboard, connects no disc blades and memory controller respectively, is used for realizing not having disc blades and is connected with memory bus between memory controller; The connecting interface of this backboard and no disc blades and memory controller comprises the signal wire of memory bus.
3. blade server as claimed in claim 1 is characterized in that: described blade server also comprises administrative unit, connects memory controller by management bus, is used for memory controller is configured and monitors.
4. as any described blade server of claim 1 to 3, it is characterized in that: described memory bus is parallel memory bus.
5. blade server as claimed in claim 1 is characterized in that: described memory bus is a universal serial bus;
Described blade server also comprises crosspoint, is connected between no disc blades and the memory controller by memory bus, is used for transmitting between no disc blades and memory controller memory access information.
6. blade server as claimed in claim 5 is characterized in that: described blade server also comprises administrative unit, connects crosspoint and memory controller by management bus, is used for crosspoint and memory controller are configured and monitor.
7. as claim 5 or 6 described blade servers, it is characterized in that: described blade server also comprises backboard, connects no disc blades and crosspoint respectively, is used for realizing not having disc blades and is connected with memory bus between crosspoint; The connecting interface of this backboard and no disc blades and crosspoint comprises the signal wire of memory bus.
8. as claim 5 or 6 described blade servers, it is characterized in that: described blade server also comprises the backboard that is integrated with crosspoint, and the connecting interface of this backboard and no disc blades and memory controller comprises the signal wire of memory bus.
9. blade server as claimed in claim 5 is characterized in that: described memory bus comprises serial attached small computer system interface SAS, serial AT computing machine optional equipment SATA, internet attached small computer system interface iSCSI, fiber optic protocols interface FCP, quick peripheral assembly interconnecting PCI-E or Infiniband bus.
10. blade server as claimed in claim 1 is characterized in that: described memory controller comprises expansion just a bunch of disks JBOD hard disk cabinet interface.
11. blade server as claimed in claim 1 is characterized in that: described storage unit is a hard disk.
12. a storage implementation method that is applied to the described blade server of claim 1 is characterized in that, may further comprise the steps:
Memory controller is divided at least one logical block with the storage unit of its management;
Memory controller is set up the logic connecting relation between logical block and the no disc blades;
No disc blades is carried out memory access by memory controller to the logical block that its logic connects.
13. storage implementation method as blade server as described in the claim 12, it is characterized in that, described method also comprises before memory controller is set up logic connecting relation between logical block and the no disc blades: no disc blades obtains device identification ID according to bus protocol, sets up logic connecting relation according to this and finishes memory access for memory controller.
14. the storage implementation method as blade server as described in claim 12 or 13 is characterized in that to comprising the blade server of administrative unit, described method also comprises: administrative unit manages it according to the duty of memory controller transmission.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101143830A CN101178634A (en) | 2006-11-08 | 2006-11-08 | Knife blade server and storage realization method therefor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2006101143830A CN101178634A (en) | 2006-11-08 | 2006-11-08 | Knife blade server and storage realization method therefor |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101178634A true CN101178634A (en) | 2008-05-14 |
Family
ID=39404904
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2006101143830A Pending CN101178634A (en) | 2006-11-08 | 2006-11-08 | Knife blade server and storage realization method therefor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101178634A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101969465A (en) * | 2010-10-13 | 2011-02-09 | 北京神州融信信息技术股份有限公司 | Cluster read-write method, apparatus and system and controller |
| CN101996111A (en) * | 2010-11-30 | 2011-03-30 | 华为技术有限公司 | Switching method, device and distributed blade server system |
| CN102006335A (en) * | 2010-10-12 | 2011-04-06 | 浪潮电子信息产业股份有限公司 | Method of storage system with multi-sever to obtain IP address of embedded memory controller |
| CN102004526A (en) * | 2010-11-02 | 2011-04-06 | 山东超越数控电子有限公司 | Vehicle fully-reinforced multi-unit server |
| CN102063268A (en) * | 2010-12-10 | 2011-05-18 | 曙光信息产业(北京)有限公司 | Control method of blade server storage equipment |
| CN102141830A (en) * | 2010-12-10 | 2011-08-03 | 曙光信息产业(北京)有限公司 | Blade server and shared memory module thereof |
| CN101540685B (en) * | 2008-06-06 | 2012-08-29 | 曙光信息产业(北京)有限公司 | PCIe shared storage blade for blade server |
| US8527565B2 (en) | 2009-03-30 | 2013-09-03 | Huawei Technologies Co., Ltd. | Selecting and reassigning a blade for a logical partition for service scheduling of a blade server |
| CN104486456A (en) * | 2014-12-11 | 2015-04-01 | 山东超越数控电子有限公司 | Method for realizing IP (internet protocol) address configuration based on I2C bus |
| WO2015081515A1 (en) * | 2013-12-04 | 2015-06-11 | 华为技术有限公司 | Data processing method, device, storage controller and equipment cabinet |
| CN105404365A (en) * | 2014-09-08 | 2016-03-16 | 广达电脑股份有限公司 | Rack system and method of operation thereof |
| CN105739930A (en) * | 2016-02-02 | 2016-07-06 | 华为技术有限公司 | Storage framework as well as initialization method, data storage method and data storage and management apparatus therefor |
| CN107547257A (en) * | 2017-07-14 | 2018-01-05 | 新华三技术有限公司 | A kind of server cluster implementation method and device |
-
2006
- 2006-11-08 CN CNA2006101143830A patent/CN101178634A/en active Pending
Cited By (17)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101540685B (en) * | 2008-06-06 | 2012-08-29 | 曙光信息产业(北京)有限公司 | PCIe shared storage blade for blade server |
| US8527565B2 (en) | 2009-03-30 | 2013-09-03 | Huawei Technologies Co., Ltd. | Selecting and reassigning a blade for a logical partition for service scheduling of a blade server |
| CN102006335A (en) * | 2010-10-12 | 2011-04-06 | 浪潮电子信息产业股份有限公司 | Method of storage system with multi-sever to obtain IP address of embedded memory controller |
| CN101969465A (en) * | 2010-10-13 | 2011-02-09 | 北京神州融信信息技术股份有限公司 | Cluster read-write method, apparatus and system and controller |
| CN102004526A (en) * | 2010-11-02 | 2011-04-06 | 山东超越数控电子有限公司 | Vehicle fully-reinforced multi-unit server |
| CN101996111A (en) * | 2010-11-30 | 2011-03-30 | 华为技术有限公司 | Switching method, device and distributed blade server system |
| CN102063268A (en) * | 2010-12-10 | 2011-05-18 | 曙光信息产业(北京)有限公司 | Control method of blade server storage equipment |
| CN102063268B (en) * | 2010-12-10 | 2013-02-13 | 曙光信息产业(北京)有限公司 | Control method of blade server storage equipment |
| CN102141830A (en) * | 2010-12-10 | 2011-08-03 | 曙光信息产业(北京)有限公司 | Blade server and shared memory module thereof |
| WO2015081515A1 (en) * | 2013-12-04 | 2015-06-11 | 华为技术有限公司 | Data processing method, device, storage controller and equipment cabinet |
| US9973351B2 (en) | 2013-12-04 | 2018-05-15 | Huawei Technologies Co., Ltd. | Data processing method and apparatus, storage controller, and cabinet |
| CN105404365A (en) * | 2014-09-08 | 2016-03-16 | 广达电脑股份有限公司 | Rack system and method of operation thereof |
| CN105404365B (en) * | 2014-09-08 | 2019-02-22 | 广达电脑股份有限公司 | Rack system and method of operating the same |
| CN104486456A (en) * | 2014-12-11 | 2015-04-01 | 山东超越数控电子有限公司 | Method for realizing IP (internet protocol) address configuration based on I2C bus |
| CN105739930A (en) * | 2016-02-02 | 2016-07-06 | 华为技术有限公司 | Storage framework as well as initialization method, data storage method and data storage and management apparatus therefor |
| CN107547257A (en) * | 2017-07-14 | 2018-01-05 | 新华三技术有限公司 | A kind of server cluster implementation method and device |
| CN107547257B (en) * | 2017-07-14 | 2021-08-24 | 新华三技术有限公司 | Server cluster implementation method and device |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101178634A (en) | Knife blade server and storage realization method therefor | |
| US7525957B2 (en) | Input/output router for storage networks | |
| US10362707B2 (en) | Reduced depth data storage assembly and rack server | |
| US6658504B1 (en) | Storage apparatus | |
| US6996642B2 (en) | Adapter, converted data storage device and method of operation of a converted data storage device | |
| US20020080575A1 (en) | Network switch-integrated high-density multi-server system | |
| US7787482B2 (en) | Independent drive enclosure blades in a blade server system with low cost high speed switch modules | |
| CN113448402B (en) | A server supporting multi-backplane cascading | |
| US20170220506A1 (en) | Modular Software Defined Storage Technology | |
| US7669045B2 (en) | System and method for aggregating shelf IDs in a fibre channel storage loop | |
| US20050182889A1 (en) | Method and apparatus for aggregating storage devices | |
| US7836204B2 (en) | Apparatus, system, and method for accessing a preferred path through a storage controller | |
| CN111475459A (en) | A 9U VPX data storage server and data processing method | |
| US7296116B2 (en) | Method and apparatus for providing high density storage | |
| CN110806989A (en) | Storage server | |
| CN104461396A (en) | Distribution type storage expansion framework based on fusion framework | |
| CN211506479U (en) | 9U VPX data storage server | |
| US12072827B2 (en) | Scaling midplane bandwidth between storage processors via network devices | |
| US20030187987A1 (en) | Storage area network with multiple pathways for command paths | |
| US20050289217A1 (en) | Low cost flexible network accessed storage architecture | |
| CN102541714B (en) | The implementation method of chip monitoring and device | |
| CN111737174A (en) | A hard disk backplane compatible with Tri-mode RAID function and its design method | |
| CN208657223U (en) | Realize the 3U high density server of centralized management | |
| US20040059901A1 (en) | Removable configuration module for storage of component configuration data | |
| CN104035524B (en) | A kind of cold storage server |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Open date: 20080514 |