CN107885672A

CN107885672A - Internal storage management system and method

Info

Publication number: CN107885672A
Application number: CN201711086105.3A
Authority: CN
Inventors: 周海江; 傅博; 郭月丰; 潘江峰; 李广斌
Original assignee: HANGZHOU ICAFE TECHNOLOGY Co Ltd
Current assignee: HANGZHOU ICAFE TECHNOLOGY Co Ltd
Priority date: 2017-11-07
Filing date: 2017-11-07
Publication date: 2018-04-06

Abstract

The application provide one kind without disk client, without disk service end, internal storage management system, method and computer-readable recording medium, wherein, applied to by least one without disk client and at least one without in disk service end network consisting system, wherein, described at least one is configured with memory modules without disk client, described at least one is configured with memory module without disk service end, the described method comprises the following steps：The memory field and the spatial information of write-back memory block configured at least one memory modules without disk client described in monitoring；According to the space-consuming of memory field and write-back memory block at least one memory modules without disk client described in spatial information distribution.

Description

Internal storage management system and method

Technical field

The application is related to network technique field, is formed more particularly to applied to by without disk client and without disk service end Internal storage management system and method in network system.

Background technology

In addition to the advantages such as hardware cost of the diskless computer system except can substantially save the storage device such as hard disk, it is even more Be easy to without being managed concentratedly in disc system without disk client, also reduce to a certain extent the maintenance of management system into This, thus diskless computer system is now widely used in the LAN environments such as Internet bar, school or even enterprise.

The user without disk client computer operation when unavoidably can be into diskless computer system without disk server in write Data, the data of these write requests are referred to as write-back, it is generally the case that one may manage very simultaneously without disk server More without disk client computer, those ceaselessly can be read or write to being sent without disk server by network in operation without disk client computer Operation requests, thus without disk server be concentrate scarce resource, if write-back can be reduced without disk client computer, nothing will be saved Disk server resource, such as the resource such as hard drive space, network traffics, CPU consumption.

The content of the invention

In view of the above the shortcomings that prior art, the purpose of the application is that provide one kind takes without disk client, without disk Business end, internal storage management system, method and computer-readable recording medium, for mitigating in diskless computer system without disk service The burden problem of device and network.

In order to achieve the above objects and other related objects, the first aspect of the application, there is provided a kind of without disk client, application In with least one network system being connected without disk service end, including：Memory modules, its memory headroom be configured as memory field and Write-back memory block；Execution module, be configurable for by network to described at least one without disk service end read data or Write-back；And it is configured as the data to be written storage when receiving write back request and arrives the write-back memory block； Distribute module, for being distributed according to the occupied space in the write-back memory block or memory field in the memory modules Deposit the space of area and write-back memory block.

The second aspect of the application, there is provided it is a kind of without disk service end, for management configuration in network system at least one Individual no disk client, including：Memory module, it is stored with the configuration processor for being read by described at least one without disk client and institute State at least one write-back write without disk client by network transmission；Distribute module, for described in general at least one without disk visitor The memory headroom at family end is configured to memory field and write-back memory block, memory field and write-back storage in the memory headroom Area is allocated according to the occupied space in the write-back memory block or memory field.

The third aspect of the application, there is provided a kind of internal storage management system, applied to by least one without disk client and extremely Few one without in disk service end network consisting system, wherein, described at least one is configured with memory modules without disk client, it is described at least One is configured with memory module without disk service end, including：Monitoring modular, for monitoring at least one internal memory mould without disk client The memory field and the spatial information of write-back memory block configured in block；Distribute module, the sky according to the monitoring module monitors Between the space-consuming of memory field and write-back memory block at least one memory modules without disk client described in information distribution.

The fourth aspect of the application, there is provided a kind of EMS memory management process, applied to by least one without disk client and extremely Few one without in disk service end network consisting system, wherein, described at least one is configured with memory modules without disk client, it is described at least One is configured with memory module without disk service end, comprises the following steps：In at least one memory modules without disk client described in monitoring The memory field of configuration and the spatial information of write-back memory block；According to described in spatial information distribution at least one without disk client The space-consuming of memory field and write-back memory block in the memory modules at end.

The 5th aspect of the application, there is provided a kind of computer-readable recording medium, be stored thereon with memory management computer Program, it is characterised in that the internal memory pipe when memory management computer program is executed by processor described in above-mentioned fourth aspect The step of reason method.

As described above, the application without disk client, without disk service end, internal storage management system, method and computer-readable Storage medium, have the advantages that：Due to the application provide embodiment without disk client during operation, from What the write-back memory block distributed in the space of memory modules act as the no disk client to a certain extent is, for example, hard The storage device of disk, since then so that without write-back to be write to the storage device of the no disk server by network In, and then network and the pressure without disk server are alleviated to a certain extent；Further, since based on user to entering without disk client The data of row write operation write-in are stored in the write-back memory block, and this allows for write-in to the write-back and deposited This partial data of storage area performs read operation again by the no disk client and can directly access memory modules when reading Write-back memory block carries out read operation to those data without again by storage device of the network access without disk service end, this Also avoid to a certain extent because network bandwidth limits the problems such as bringing reading or writing speed slow, and then also to a certain extent Alleviate the burden without disk server and network in diskless computer system.

Brief description of the drawings

Fig. 1 is shown as including no disk service end management and the network system architecture schematic diagram without disk client.

Fig. 2 is shown as application architecture schematic diagram of the application without disk client.

Fig. 3 is shown as application architecture schematic diagram of the application without disk service end.

Fig. 4 is shown as the configuration diagram of the application internal storage management system.

Fig. 5 is shown as the schematic flow sheet of the EMS memory management process of the application in one embodiment.

Fig. 6 is shown as the schematic flow sheet of the EMS memory management process of the application in another embodiment.

Embodiment

Presently filed embodiment is illustrated by particular specific embodiment below, those skilled in the art can be by this explanation Content disclosed by book understands other advantages and effect of the application easily.

In described below, refer to the attached drawing, accompanying drawing describes some embodiments of the application.It should be appreciated that also it can be used Other embodiment, and module composition and operational can be carried out in the case of without departing substantially from spirit and scope of the present disclosure Change.Following detailed description should not be considered limiting, and the scope of embodiments herein is only by announcing Claims of patent are limited.Term used herein is merely to describe specific embodiment, and be not intended to limit this Application.Although the grade of term first, second is used for describing various elements herein in some instances, these elements should not When being limited by these terms.These terms are only used for making a distinction an element with another element.Furthermore as herein Used in, singulative " one ", "one" and "the" are intended to also include plural form, unless there is opposite finger in context Show.It will be further understood that term "comprising", " comprising " show described feature, step, operation, element, component, item be present Mesh, species, and/or group, but be not excluded for other one or more features, step, operation, element, component, project, species and/ Or presence, appearance or the addition of group.

The application provide it is a kind of without disk client, applied to at least one network system being connected without disk service end, institute The communication network for stating network system can be internet, one or more in-house networks, LAN (LAN), wide area network (WLAN), deposit Store up LAN (SAN) etc., or its is appropriately combined.Referring to Fig. 1, it is shown as including no disk service end management and without disk client 21 network system architecture schematic diagram, one as depicted without disk service end 11, by above-mentioned communication network, management is more simultaneously Individual no disk client 21, the no disk client 21 are also referred to as non-disk workstation, and the no disk client 21 passes through net Network is connected with without disk service end 11, and no disk client 21 is from without the downloading guides program of disk service end 11, configuration file, kernel and mirror As file etc., with the start-up operation system without disk client 21 of realization.For example the operating system is when being (SuSE) Linux OS, Wherein, the bootstrap is the boot loader in (SuSE) Linux OS, and the kernel is in (SuSE) Linux OS Kernel, the image file are that the initrd compressed files of init files, the configuration file are included in (SuSE) Linux OS Title including kernel, kernel particular location, the title of image file, the particular location of image file, and start Kernel parameter (for example, IP arrange parameters, start-up mode parameter, hot-plug equipment parameter).

The no disk client 21 be without disk (not including hard disk) personal computer, notebook computer and it is other arbitrarily The equipment or system of non-disk startup can be realized.The no disk service end 11 can be personal computer, the webserver (including Storage server and application server etc.), it can also be any other applicable computer.In certain embodiments, the nothing Disk service end 11 can be dynamic host allocation protocol (Dynamic Host Configuration Protocol, abbreviation DHCP) server, no disk service end 11 is in Pre-boot Execution Environment (Preboot Execution Environment, abbreviation PXE with being communicatively coupled without disk client 21 under).

Referring to Fig. 2, application architecture schematic diagram of the application without disk client is shown as, as illustrated, the no disk visitor Family end 21 includes：Memory modules 210, execution module 211, and distribute module 212.

The memory headroom of the memory modules 210 is configured as memory field and write-back memory block；In the implementation of reality In mode, the memory modules 210 are the internal memory (Memory) of the no disk client 21, also referred to as built-in storage, are used for The temporarily operational data in storage central processing unit (Central Processing Unit, abbreviation CPU), and from described It is, for example, the data that the external memory storages such as the storage device of hard disk exchange without disk server, for example draws from being downloaded without disk service end 11 Helical pitch sequence, configuration file, kernel and image file etc. are with the data of start-up operation system of realization without disk client 21 etc..It is described Memory field is used to store from bootstrap, kernel, image file and configuration file without the down operation system of disk service end 11 etc., Also for storage in corresponding program and file run without disk client 21 needed for some softwares etc..The write-back memory block For storing based on user to the data without the progress write operation write-in of disk client 21, in other words, run without disk client 21 During, what the write-back memory block act as the no disk client 21 to a certain extent is, for example, the storage of hard disk Equipment.

For example, the memory headroom of the memory modules 210 is 8G, based on some configuration strategies, the internal memory for 8G is empty Between be configured as example, 4G memory field and 4G write-back memory block.Certainly, in different situations, the internal memory mould The memory headroom of block 210 can be configured flexibly according to the situation of reality, for example the memory headroom for 8G is configured as example The write-back memory block of memory field and 6G for 2G；Or the memory headroom for 8G is configured as example, 6G internal memory Area and 2G write-back memory block etc., the not exhaustion herein of all so examples.

It should be strongly noted that different based on the hardware configuration without disk client 21, the internal memory of its memory modules 210 is empty Between may be 2G, 4G, 16G or 32G or even more, in the following description, in order to concisely illustrate the thought of the application And principle, temporarily illustrated so that the memory headroom of the memory modules 210 without disk client 21 is 8G as an example.In addition, hereinafter institute Either the size in space and capacity are generally described the data stated with modes such as 1G or 2G, and those skilled in the art could be aware that , described 1G refers to that 1024M, described 2G refer to 2048M, and such it will not go into details.

The execution module 211 be configurable for by network to described at least one without disk service end 11 read data or Write-back；And it is configured as the data to be written storage when receiving write back request and arrives the write-back memory block； In the embodiment of reality, the execution module 211 is, for example, central processing unit without disk service end 11 or by described Combination of one or more programs or hardware and software performed by central processing unit etc..The no disk client 21 is starting Stage needs to carry out read operation, i.e., by network from described at least one without the downloading guides program of disk service end 11, configuration file, interior Core and image file etc., with the start-up operation system without disk client 21 of realization, the no disk client 21 is based on user's Operation is set when carrying out write operation, i.e., by network to the described at least one storage device write-back number without disk service end 11 According to the execution module 211 in the application is additionally configured to, and when receiving write back request without disk client 21, first will The data of write-in are written to the write-back memory block configured in the memory modules 210, when the operation for receiving user again Or the data stored in the write-back memory block are written back to by an at least nothing by network based on default strategy In the storage device of disk service end 11.

The distribute module 212 is used for according to the occupied space in the write-back memory block or memory field to distribute State the space of memory field and write-back memory block in memory modules 210.In the embodiment of reality, the distribute module 212 be, for example, central processing unit without disk client 21 or by one or more programs performed by the central processing unit Or combination of hardware and software etc..

The distribute module 212, which distributes the space of memory field and write-back memory block in the memory modules 210, to be included Two kinds of situations, a kind of is that the space in memory modules 210 is allocated first, and another kind is, based on the allocated internal memory The space of memory field and write-back memory block in module 210 is reallocated, that is, is redistributed.

In one embodiment in the first scenario, the distribute module 212 is used for occupied according to the memory field Space to distribute the space of memory field and write-back memory block in the memory modules 210.Such as the no disk client The memory headroom of memory modules 210 at end 21 is 8G, wherein, based on the monitoring of the no disk client 21 to being currently running, The program of its memory field operation has occupied 2.5G space, to ensure the internally stored program that may increase operation, then for institute Then need to consider that the memory field is given in the space that memory field is enough, such as distribution 5G space when stating memory field allocation space, its Remaining 3G space is as write-back memory block.

In an embodiment in the latter case, the distribute module 212 is based on the allocated memory modules 210 In memory field and write-back memory block space carry out space allocation, that is, redistribute.Such as when the no disk client The memory headroom of 21 memory modules 210 is 8G, and it is assigned with 4G spaces as memory field, 4G in addition for 8G memory headroom Space is as write-back memory block, and based on the monitoring of the no disk client 21 to being currently running, it is whole interior that discovery writes it Deposit the program run in area and only taken up 1.5G space, and estimate the program that it runs in certain period of time do not have it is too big Change, then can carry out space allocation to the space of the write-back memory block, for example configure again from the memory field Go out 1G and distribute to the write-back memory block, that is, the space configuration in the memory modules 210 redistributed is 3G memory field With 5G write-back memory block；Or the monitoring based on the no disk client 21 to being currently running, monitor in described When memory field remaining space is less than 1G in storing module 210, to ensure the program that may subsequently run to the further of memory headroom Demand then needs to redistribute the write-back memory block in the memory modules, it is necessary to extend the space of the memory field Space-consuming in 210, for example configure 2G from the write-back memory block again and distribute to the memory field, i.e., divide again The memory field and 2G write-back memory block that space configuration in the memory modules 210 matched somebody with somebody is 6G.

In the implementation process of above-mentioned two situations, it can be based on empty to write-back memory block spatial information and memory field Between the threshold value (minimum spatial value) of information set to realize dynamic and non-dynamic distribution, such as, such as the no disk client The memory headroom of memory modules 210 at end 21 be 8G, and it is that 8G memory headroom is assigned with 4G spaces as memory field, in addition 4G spaces as write-back memory block, when the distribution of the gross space of internal memory must not be arranged to be less than 3G when, even if based on pair The monitoring for the no disk client 21 being currently running finds that writing the program run in its whole memory field has only taken up 1.5G's Space, and estimate the program that it runs in certain period of time and do not have too big change, but also will not be to the write-back The space of memory block carries out larger space allocation, for example 2G will not be marked off from the memory field and distributes to the write-back number According to memory block, and then the memory field of the space division in the memory modules 210 for ensuring to redistribute will not be less than 3G space.

Due to the application without disk client 21 during operation, distributed from the space of memory modules 210 return The storage device that data storage area act as for example, hard disk of the no disk client 21 to a certain extent is write, so that So that without write-back is write in the storage device of the no disk server by network, and then alleviate to a certain extent Network and the pressure without disk server；Further, since it is to the data that write operation write-in is carried out without disk client 21 based on user It is stored in the write-back memory block, this allows for write-in to this partial data of the write-back memory block by institute The write-back memory block pair of memory modules 210 can directly be accessed by stating when no disk client 21 performs read operation again and read Those data carry out read operation without by storage device of the network access without disk service end 11, this also keeps away to a certain extent again Exempt from because network bandwidth limits the problems such as bringing reading or writing speed slow.

In some embodiments of the application, the distribute module 212 is additionally operable to monitor to return in the memory modules 210 Write and take the write-back being stored in by network storage to the no disk when data storage area remaining space is less than certain threshold value End 11 be engaged in discharge the space-consuming of the write-back memory block.Such as the memory modules 210 when the no disk client 21 Memory headroom be 8G, it is assigned with 4G spaces as memory field for 8G memory headroom, and 4G spaces in addition are as write-back number According to memory block, the default threshold value is, for example, 1024M or 512M, when the distribute module 212 also monitors the write-back When the remaining space of data storage area is less than 1024M or 512M, the data to ensure subsequently to write can have enough spaces The write-back memory block is written into, then needs the data for having been written into the write-back memory block being transferred to by network Without in disk server, and then the space-consuming of the write-back memory block is discharged, it be 4G again to make it, and then is met subsequently Write demand of the data to memory space.This implementation is unnecessary to redistribute memory headroom, i.e., is stored in write-back During the remaining space deficiency in area and the space resources of memory field need not be seized, simply to write-back memory block according to default Rule carries out data conversion storage and empties the write-back memory block.

In some embodiments of the application, the distribute module 212 can dynamically adjust depositing in memory modules 210 Space is stored up, i.e., described distribute module 212 is actually accounted for for monitoring the memory field or the write-back memory block in real time Space is with the memory field in memory modules described in dynamically distributes 210 and the space of write-back memory block.

In some embodiments of the application, the distribute module 212 is additionally operable to monitor to return in the memory modules 210 Write and redistribute the write-back memory block in the memory modules 210 when data storage area remaining space is less than certain threshold value In space-consuming.Such as when the memory headroom of the memory modules 210 of the no disk client 21 is 8G, it is empty for 8G internal memory Between be assigned with 4G spaces as memory field, other 4G spaces are as write-back memory block, based on to described in being currently running Monitoring without disk client 21, it is found that remaining space has been less than certain threshold value in its write-back memory block, for example, it is described default Threshold value be, for example, 1024M or 512M, when the distribute module 212 also monitors that the residue of the write-back memory block is empty Between less than 1024M or 512M when, based on the monitoring of the no disk client 21 to being currently running, find its whole memory field The program of middle operation has only taken up 1.5G space, and estimates the program that it runs in certain period of time and do not have too big change Change, then can carry out space allocation to the space of the write-back memory block, for example from the memory field configure 1G again The write-back memory block is distributed to, that is, the space configuration in the memory modules 210 redistributed is 3G memory field and 5G Write-back memory block.

In some embodiments of the application, the distribute module 212 is additionally operable to monitor the write-back of execution module 211 please Frequency is sought to redistribute space-consuming of the write-back memory block in the memory modules 210.Such as when the nothing The memory headroom of the memory modules 210 of disk client 21 is 8G, and it is assigned with 4G spaces as memory field for 8G memory headroom, Other 4G spaces, based on the monitoring of the no disk client 21 to being currently running, monitor as write-back memory block The execution module 211 continually to the write-back memory block write data, it is described frequently refer to such as one minute or The frequency of write-back is more than 3 times or 2 times in half a minute, then system is thought because frequently write operation may result in described The spatial content deficiency of write-back memory block in storing module 210, and then need to redistribute the write-back memory block and exist Space-consuming in the memory modules 210, for example from the memory field configure 1G again and distribute to the write-back and deposit Storage area, that is, the memory field and 5G write-back memory block that the space configuration in the memory modules 210 redistributed is 3G.For The description of the write operation frequency is not limited thereto, and frequency can be set according to the situation of reality in different application scenarios and example Secondary threshold value, will not be described here.

In some embodiments of the application, the distribute module 212 is additionally operable to monitor the current of the execution module 211 Write-back amount is to redistribute space-consuming of the write-back memory block in the memory modules 210.

The current write-back amount such as be current time single write-back amount, such as the data volume that single writes is During 256M, then system think due to single write-in data volume it is larger, may subsequently still suffer from larger data volume be written into And the spatial content deficiency of write-back memory block in the memory modules 210 can be caused, and then need to redistribute described return Write space-consuming of the data storage area in the memory modules 210.

The current write-back amount is such as the write-back quantity in current time or recent write-back amount, that is, is worked as Write-back amount in the previous period, when write-back amount is 256M in for example, current one minute, then system thinks Because lasting write-in data volume is larger, it may subsequently still suffer from larger data volume and be written into and then the internal memory mould can be caused The spatial content deficiency of write-back memory block in block 210, and then need to redistribute the write-back memory block described Space-consuming in memory modules 210.

The current write-back amount is such as continuous write-back amount several times, write-back for example, twice in succession Measure for 256M when, then system thinks that data volume is larger due to being continuously written into, and may subsequently still suffer from larger data volume and be write Enter and then the spatial content deficiency of write-back memory block in the memory modules 210 can be caused, and then need to redistribute institute State space-consuming of the write-back memory block in the memory modules 210.

Such as when the memory headroom of the memory modules 210 of the no disk client 21 is 8G, it divides for 8G memory headroom 4G spaces are matched somebody with somebody as memory field, 4G spaces in addition are as write-back memory block, based on the no disk to being currently running The monitoring of client 21, when monitoring that the current write-back amount of the execution module 211 is more than a certain threshold value, this threshold value example Such as be 256M, then system think due to write-in data volume it is larger, may subsequently still suffer from larger data volume be written into and then The spatial content deficiency of write-back memory block in the memory modules 210 can be caused, and then need to redistribute the write-back Space-consuming of the data storage area in the memory modules 210, for example configure 1G from the memory field again and distribute to institute Write-back memory block is stated, that is, the space configuration in the memory modules 210 redistributed is 3G memory field and 5G write-back number According to memory block.Description for the data volume size of write operation write-in is not limited thereto, in different application scenarios and example In the threshold value of writing can be set according to the situation of reality, such as 128M, 64M etc. will not be described here.

In some embodiments of the application, the distribute module 212 is additionally operable to monitor in the memory modules 210 interior Deposit and space-consuming of the memory field in the memory modules 210 is redistributed when area's remaining space is less than certain threshold value.Example Such as when the memory headroom of the memory modules 210 of the no disk client 21 is 8G, it is assigned with 4G spaces for 8G memory headroom As memory field, 4G spaces in addition are as write-back memory block, based on the no disk client 21 to being currently running Monitoring, when monitoring that memory field remaining space is less than 1G in the memory modules 210, to ensure the program pair that may subsequently run The further demand of memory headroom then needs to redistribute the write-back storage, it is necessary to extend the space of the memory field Space-consuming of the area in the memory modules 210, for example configure 2G from the write-back memory block again and distribute to institute Memory field is stated, that is, the memory field and 2G write-back memory block that the space configuration in the memory modules 210 redistributed is 6G. Description for memory field remaining space deficiency situation is not limited thereto, can foundation in different application scenarios and example Actual situation sets the threshold value of remaining space, will not be described here.

In some embodiments of the application, the distribute module 212 is additionally operable to monitor in the memory modules 210 interior Deposit the write-back being stored in when area's remaining space is less than certain threshold value by network storage to the no disk service end 11, and Redistribute space-consuming of the memory field in the memory modules 210.Such as the internal memory when the no disk client 21 The memory headroom of module 210 is 8G, and it is assigned with 4G spaces as memory field, 4G spaces conduct in addition for 8G memory headroom Write-back memory block, based on the monitoring of the no disk client 21 to being currently running, monitor in the memory modules 210 When memory field remaining space is less than 1G, to ensure the program that may subsequently run to the further demand of memory headroom, it is necessary to expand The space of memory field described in exhibition, then need to redistribute occupancy of the write-back memory block in the memory modules 210 Space, if now because write-in data cause it not have the space of redundancy to distribute in described in the write-back memory block Area is deposited, then the data for needing to store in the write-back memory block, to the no disk service end 11, are changed by network storage Yan Zhi, the data in write-back memory block are carried out with unloading and empties the write-back memory block, so as to for its from New allocation space, for example configure 2G from the write-back memory block again and distribute to the memory field, that is, redistribute The memory field and 2G write-back memory block that space configuration in memory modules 210 is 6G.For memory field remaining space not The description of sufficient situation is not limited thereto, and can be set in different application scenarios and example according to the situation of reality remaining empty Between threshold value, will not be described here.

The application's can dynamically adjust the sky of memory field and write-back memory block in memory modules without disk client Between so that without the disk client actual occupancy situation according to internal memory in operation, the write-back memory block of dynamic adjustment in real time Space, and then ensure no disk client will not because its memory modules by write-back take bring low memory, while Meet and be locally stored and reading the demand of write-back without disk client, certainly, also cause without write-back is passed through Network is write in the storage device of the no disk server, alleviates network and the pressure without disk server to a certain extent.

The application also provides one kind without disk service end, at least one no disk client of the management configuration in network system End, the communication network of the network system can be internet, one or more in-house networks, LAN (LAN), wide area network (WLAN), storage area network (SAN) etc., or its is appropriately combined.It is multiple to be managed referring to Fig. 1, being shown as including no disk service end 11 Reason and the network system architecture schematic diagram without disk client 21, one as depicted without disk service end 11 by above-mentioned communication Network manages multiple no disk clients 21 simultaneously, and the no disk client 21 is also referred to as non-disk workstation, the no disk Client 21 is connected by network with without disk service end 11, no disk client 21 from without the downloading guides program of disk service end 11, match somebody with somebody File, kernel and image file etc. are put, with the start-up operation system without disk client 21 of realization.For example the operating system is During (SuSE) Linux OS, wherein, the bootstrap is the boot loader in (SuSE) Linux OS, and the kernel is Kernel in (SuSE) Linux OS, the image file are that the initrd comprising init files compresses in (SuSE) Linux OS File, the configuration file include kernel title, kernel particular location, the title of image file, image file Particular location, and start kernel parameter (for example, IP arrange parameters, start-up mode parameter, hot-plug equipment parameter).

Referring to Fig. 3, application architecture schematic diagram of the application without disk service end is shown as, as illustrated, the no disk clothes Business end 11 includes memory module 111 and distribute module 112.

The memory module 111 is stored with for by least one configuration processor read without disk client 21 and described At least one write-back write without disk client 21 by network transmission；Such as the memory module 111 is stored with for quilt Described at least one without the configuration processor that disk client 21 is read include bootstrap, configuration file, kernel and image file etc. with The program and file of the start-up operation system without disk client 21 of realization, also include such as user and running certain without disk client 21 Corresponding program and file needed for a little softwares.The memory module 111 is, for example, high-speed random access memory, and can also be wrapped Nonvolatile memory is included, such as one or more disk storage equipments, flash memory device or other nonvolatile solid state storages are set It is standby.In certain embodiments, memory can also include the memory away from one or more processors, such as via RF circuits Or outside port and the network attached storage of communication network access.

The distribute module 112 is used at least one memory headroom without disk client 21 being configured to memory field and returned Write data storage area, memory field and write-back memory block are according to the write-back memory block or interior in the memory headroom Deposit what the occupied space in area was allocated.In the embodiment of reality, the distribute module 112 is, for example, without disk service end 11 central processing unit is either by the one or more programs or the knot of hardware and software performed by the central processing unit Close etc..

In certain embodiments, the memory headroom of the no disk client 21 is configured as memory field and write-back storage Area；In the embodiment of reality, the memory headroom is the internal memory (Memory) of the no disk client 21, also referred to as interior Memory, for temporarily depositing the operational data in central processing unit (Central Processing Unit, abbreviation CPU), with And from the no disk server be, for example, the data that the external memory storages such as the storage device of hard disk exchange, such as from without disk service 11 downloading guides programs, configuration file, kernel and image file etc. are held to realize the start-up operation system without disk client 21 Data etc..The memory field be used for store from the bootstrap without the down operation system of disk service end 11, kernel, image file and Configuration file etc., also for storage in corresponding program and file run without disk client 21 needed for some softwares etc..Described time Write data storage area be used for store based on user to without disk client 21 progress write operation write-in data, in other words, without disk During client 21 is run, the write-back memory block act as the no disk client 21 for example to a certain extent For the storage device of hard disk.

For example, the memory headroom of the memory headroom is 8G, and based on some configuration strategies, the memory headroom quilt for 8G It is configured to for example, 4G memory field and 4G write-back memory block.Certainly, in different situations, the memory headroom Memory headroom can be configured flexibly according to the situation of reality, for example the memory headroom for 8G is configured as the interior of for example, 2G Deposit area and 6G write-back memory block；Or the memory headroom for 8G is configured as example, 6G memory field and 2G Write-back memory block etc., the not exhaustion herein of all so examples.

The distribute module 112, which distributes the space of memory field and write-back memory block in the memory headroom, includes two kinds Situation, a kind of is that the space in memory headroom is allocated first, and another kind is, based in the allocated memory headroom The space of memory field and write-back memory block is reallocated, that is, is redistributed.

In one embodiment in the first scenario, the distribute module 112 is used for occupied according to the memory field Space to distribute the space of memory field and write-back memory block in the memory headroom.Such as when the no disk client 21 The memory headroom of memory headroom be 8G, wherein, based on the monitoring of the no disk client 21 to being currently running, its memory field The program of operation has occupied 2.5G space, may increase the internally stored program of operation to ensure, then for the memory field Then need to consider that the memory field, remaining 3G sky are given in the space that memory field is enough, such as distribution 5G space during allocation space Between be used as write-back memory block.

In an embodiment in the latter case, the distribute module 112 is based in the allocated memory headroom Memory field and write-back memory block space carry out space allocation, that is, redistribute.Such as when the no disk client 21 The memory headroom of memory headroom be 8G, it is assigned with 4G spaces as memory field, 4G spaces in addition for 8G memory headroom As write-back memory block, based on the monitoring of the no disk client 21 to being currently running, its whole memory field is write in discovery The program of middle operation has only taken up 1.5G space, and estimates the program that it runs in certain period of time and do not have too big change Change, then can carry out space allocation to the space of the write-back memory block, for example from the memory field configure 1G again Distribute to the write-back memory block, that is, memory field that the space configuration in the memory headroom redistributed is 3G and 5G's returns Write data storage area；Or the monitoring based on the no disk client 21 to being currently running, monitor in the memory headroom When memory field remaining space is less than 1G, to ensure the program that may subsequently run to the further demand of memory headroom, it is necessary to expand The space of memory field described in exhibition, then need to redistribute occupancy sky of the write-back memory block in the memory headroom Between, for example configure 2G from the write-back memory block again and distribute to the memory field, that is, the memory headroom redistributed In space configuration be 6G memory field and 2G write-back memory block.

In certain embodiments, the no disk service end 11 can with management configuration in network system multiple no disk clients 21, for monitoring the situation of memory headroom in described multiple no disk clients 21, and can unify to those without disk client 21 Internal memory carry out unified configuration, described configuration refers to that the mode of batch operation will be respective in the multiple no disk client 21 Memory headroom is configured to memory field and write-back memory block.

In certain embodiments, the no disk service end 11 can with management configuration in network system multiple no disk clients 21, for monitoring the situation of memory headroom in described multiple no disk clients 21, and can be pointedly to wherein there is a nothing The internal memory of disk client 21 is configured, and described configuration refers to memory headroom in one no disk client 21 being configured to Memory field and write-back memory block.

Due to the application without disk service end 11 during the operation of at least one no disk client 21 is supervised, from reality Serve as to a certain extent the write-back memory block distributed in the space of described at least one no memory headroom of disk client 21 The no disk client 21 be, for example, hard disk storage device, since then so that without write-back is passed through into network Write in the storage device of the no disk server, and then alleviate network and the pressure without disk server to a certain extent；Separately Outside, due to being stored in based on user to the data that write operation write-in is carried out without disk client 21 in the write-back memory block , this allows for write-in and is performed reading behaviour again by the no disk client 21 to this partial data of the write-back memory block The write-back memory block that memory headroom can be directly accessed when making and reading carries out read operation to those data without leading to again Cross storage device of the network access without disk service end 11, this also avoid to a certain extent because network bandwidth limitation bring reading or The problems such as person's writing speed is slow.

In some embodiments of the application, the distribute module 112 is additionally operable to monitor at least one no disk visitor The write-back being stored in is passed through when write-back memory block remaining space is less than certain threshold value in the memory headroom at family end 21 Network storage is to the memory module 111 to discharge the space-consuming of the write-back memory block.Such as the no disk visitor The memory headroom of the memory headroom at family end 21 is 8G, and it is assigned with 4G spaces as memory field for 8G memory headroom, in addition 4G spaces are, for example, 1024M or 512M as write-back memory block, the default threshold value, when the distribute module 112 When also monitoring the remaining space of the write-back memory block less than 1024M or 512M, to ensure the data subsequently write There can be enough spaces to be written into the write-back memory block, then need that the number of the write-back memory block will be had been written into It is transferred to according to by network in no disk server, and then discharges the space-consuming of the write-back memory block, makes it be again 4G, and then meet demand of the follow-up write-in data to memory space.This implementation is unnecessary to redistribute memory headroom, I.e. in the remaining space deficiency of write-back memory block and the space resources of memory field need not be seized, simply to write-back Memory block carries out data conversion storage according to default rule and empties the write-back memory block.

In some embodiments of the application, the distribute module 112 can dynamically adjust the storage in memory modules Space, i.e., described distribute module 112 are actually taken for monitoring the memory field or the write-back memory block in real time Space with the memory field in memory headroom described in dynamically distributes and the space of write-back memory block.

In some embodiments of the application, the distribute module 112 is additionally operable to monitor at least one no disk visitor The write-back is redistributed when write-back memory block remaining space is less than certain threshold value in the memory headroom at family end 21 to deposit Storage area is in the space-consuming in the memory headroom.Such as when the memory headroom of the memory headroom of the no disk client 21 is 8G, it is assigned with 4G spaces for 8G memory headroom and is based on as memory field, 4G spaces in addition as write-back memory block To the monitoring for the no disk client 21 being currently running, it is found that remaining space has been less than certain threshold in its write-back memory block Value, for example the default threshold value is, for example, 1024M or 512M, when the distribute module 112 also monitors the write-back number During according to the remaining space of memory block less than 1024M or 512M, based on the monitoring of the no disk client 21 to being currently running, It was found that the program run in its whole memory field has only taken up 1.5G space, and estimate its journey run in certain period of time Sequence does not have too big change, then can carry out space allocation to the space of the write-back memory block, such as again from described 1G is configured in memory field and distributes to the write-back memory block, that is, the space configuration in the memory headroom redistributed is 3G Memory field and 5G write-back memory block.

In some embodiments of the application, the distribute module 112 is additionally operable to described in monitoring at least one without disk client 21 write back request frequency is to redistribute the space-consuming of the write-back memory block in its memory headroom.Such as when the nothing The memory headroom of the memory headroom of disk client 21 is 8G, and it is assigned with 4G spaces as memory field for 8G memory headroom, separately Outer 4G spaces, based on the monitoring of the no disk client 21 to being currently running, monitor institute as write-back memory block State no disk client 21 and continually write data to the write-back memory block, it is described frequently to refer to such as one minute or half The frequency of write-back is more than 3 times or 2 times in minute, then system is thought because frequently write operation may result in the internal memory The spatial content deficiency of write-back memory block in space, and then need to redistribute the write-back memory block described interior Deposit the space-consuming in space, for example configure 1G from the memory field again and distribute to the write-back memory block, i.e., it is heavy The memory field and 5G write-back memory block that space configuration in newly assigned memory headroom is 3G.For the write operation frequency Description is not limited thereto, and can set the threshold value of the frequency according to the situation of reality in different application scenarios and example, It will not go into details for this.

In some embodiments of the application, the distribute module 112 is additionally operable to described in monitoring at least one without disk client 21 current write-back amount is to redistribute the space-consuming of the write-back memory block in its memory headroom.

Such as when the memory headroom of the memory headroom of the no disk client 21 is 8G, it is assigned with for 8G memory headroom 4G spaces are as memory field, and 4G spaces in addition are as write-back memory block, based on the no disk client to being currently running The monitoring at end 21, when monitoring that the current write-back amount of the no disk client 21 is more than a certain threshold value, this threshold value is for example For 256M, then system is thought because write-in data volume is larger, may subsequently be still suffered from larger data volume and is written into and then meeting Cause the spatial content deficiency of write-back memory block in the memory headroom, and then need to redistribute the write-back and deposit Space-consuming of the storage area in the memory headroom, for example configure 1G from the memory field again and distribute to the write-back Memory block, that is, the memory field and 5G write-back memory block that the space configuration in the memory headroom redistributed is 3G.For The description of the data volume size of write operation write-in is not limited thereto, can be according to reality in different application scenarios and example Situation the threshold value of writing is set, such as 128M, 64M etc. will not be described here.

In some embodiments of the application, the distribute module 112 is additionally operable to monitor described at least one without disk client The memory field is redistributed in the memory headroom when memory field remaining space is less than certain threshold value in the memory headroom at end 21 In space-consuming.Such as when the memory headroom of the memory headroom of the no disk client 21 is 8G, it is 8G memory headroom 4G spaces are assigned with as memory field, 4G spaces in addition are as write-back memory block, based on the nothing to being currently running The monitoring of disk client 21, subsequently may fortune to ensure when monitoring that memory field remaining space is less than 1G in the memory headroom Capable program, it is necessary to extend the space of the memory field, then needs to redistribute described return to the further demand of memory headroom Space-consuming of the data storage area in the memory headroom is write, for example configures 2G points from the write-back memory block again Memory field described in dispensing, that is, the memory field and 2G write-back storage that the space configuration in the memory headroom redistributed is 6G Area.Description for memory field remaining space deficiency situation is not limited thereto, can be with different application scenarios and example The threshold value of remaining space is set according to actual situation, will not be described here.

In some embodiments of the application, the distribute module 112 is additionally operable to monitor described at least one without disk visitor The write-back being stored in is passed through into network storage when memory field remaining space is less than certain threshold value in the memory headroom at family end 21 To the memory module 111, and redistribute space-consuming of the memory field in the memory headroom.Such as when the nothing The memory headroom of the memory headroom of disk client 21 is 8G, and it is assigned with 4G spaces as memory field for 8G memory headroom, separately Outer 4G spaces, based on the monitoring of the no disk client 21 to being currently running, monitor institute as write-back memory block When stating that memory field remaining space is less than 1G in memory headroom, be ensure subsequently may operation program to the further of memory headroom Demand then needs to redistribute the write-back memory block in the memory headroom, it is necessary to extend the space of the memory field In space-consuming, if now in the write-back memory block because write-in data cause it there is no the space of redundancy to distribute To the memory field, then the data for needing to store in the write-back memory block pass through network storage to the no disk service End 11, in other words, the data in write-back memory block are carried out with unloading and empties the write-back memory block, so as to It is it from new allocation space, for example configures 2G from the write-back memory block again and distribute to the memory field, i.e., again The memory field and 2G write-back memory block that space configuration in the memory headroom of distribution is 6G.For memory field remaining space The description of insufficient situation is not limited thereto, and can be set in different application scenarios and example according to the situation of reality remaining The threshold value in space, will not be described here.

The application's can dynamically adjust memory field and write-back memory block in memory modules without disk service end 11 Space so that without disk client 21 in operation according to internal memory actual occupancy situation, in real time dynamic adjustment write-back deposit The space of storage area, and then ensure that no disk client 21 will not bring low memory because its memory modules is taken by write-back, Also meet simultaneously and be locally stored and reading the demand of write-back without disk client 21, certainly, also causing need not be by write-back Data are write by network in the storage device of the no disk server, alleviate network to a certain extent and without disk server Pressure.

The application also provides a kind of internal storage management system, is taken applied to by least one without disk client and at least one without disk In end network consisting system of being engaged in, wherein, described at least one is configured with memory modules without disk client, and described at least one without disk service End is configured with memory module.The communication network of the network system can be internet, one or more in-house networks, LAN (LAN), wide area network (WLAN), storage area network (SAN) etc., or its is appropriately combined.Again referring to Fig. 1, being shown as including nothing Disk service end 11 manages and the network system architecture schematic diagram without disk client 21, one as depicted without disk service end 11 by Multiple no disk clients 21 are managed by above-mentioned communication network simultaneously, the no disk client 21 is also referred to as no disk work Stand, the no disk client 21 is connected by network with without disk service end 11, and no disk client 21 is downloaded from without disk service end 11 Bootstrap, configuration file, kernel and image file etc., with the start-up operation system without disk client 21 of realization.It is such as described When operating system is (SuSE) Linux OS, wherein, the bootstrap be (SuSE) Linux OS in boot loader, institute Kernel is stated as the kernel in (SuSE) Linux OS, the image file is comprising init files in (SuSE) Linux OS Initrd compressed files, the configuration file include kernel title, kernel particular location, the title of image file, The particular location of image file, and start kernel parameter (for example, IP arrange parameters, start-up mode parameter, hot plug are set Standby parameter).

The no disk client 21 be without disk (not including hard disk) personal computer, notebook computer and it is other arbitrarily The equipment or system of non-disk startup can be realized.The no disk service end 11 can be personal computer, the webserver (including Storage server and application server etc.), it can also be any other applicable computer.In the embodiment of reality, institute The internal memory (Memory) that memory modules are the no disk client 21, also referred to as built-in storage are stated, for temporarily depositing center Operational data in processor (Central Processing Unit, abbreviation CPU), and from the no disk server example The data such as exchanged for the storage device external memory storage of hard disk, such as from without the downloading guides program of disk service end 11, configuration File, kernel and image file etc. are with the data of start-up operation system of realization without disk client 21 etc..

The no disk service end 11 can be dynamic host allocation protocol (Dynamic Host Configuration Protocol, abbreviation DHCP) server, no disk service end 11 is in Pre-boot Execution Environment (Preboot Execution Environment, abbreviation PXE) under with being communicatively coupled without disk client 21.In the embodiment of reality, the storage Module be stored with for by least one configuration processor read without disk client 21 and described at least one without disk client 21 The write-back write by network transmission；Such as the memory module is stored with and is used for by described at least one without disk client 21 The configuration processor of reading includes bootstrap, configuration file, kernel and image file etc. to realize the startup without disk client 21 The program and file of operating system, also include such as user run without disk client 21 corresponding program needed for some softwares and File.The memory module is, for example, high-speed random access memory, and may also include nonvolatile memory, such as one Or multiple disk storage equipments, flash memory device or other non-volatile solid-state memory devices.In certain embodiments, memory is also The memory away from one or more processors can be included, such as via RF circuits or outside port and communication network access Network attached storage.

Referring to Fig. 4, the configuration diagram of the application internal storage management system is shown as, as illustrated, the memory management System 3 includes monitoring modular 30, distribute module 31, for purposes of illustration only, also including being configured at no disk client in the diagram drawn The memory modules 32 at end and the memory module 33 for being configured at no disk service end.

The monitoring modular 30 is used to monitor the internal memory configured at least one memory modules 32 without disk client 21 Area and the spatial information of write-back memory block；In certain embodiments, it is, for example, to be taken without disk that the monitoring modular 30, which is, for example, The central processing unit at business end 11 is either by the one or more programs or hardware and software performed by the central processing unit Combination etc..

The memory headroom of the memory modules 32 is configured as memory field and write-back memory block；The memory field is used for Storage is also used for storing from bootstrap, kernel, image file and configuration file without the down operation system of disk service end 11 etc. Corresponding program and file needed for some softwares etc. are being run without disk client 21.The write-back memory block is used to store base In user to without disk client 21 carry out write operation write-in data, in other words, without disk client 21 run during, institute State the storage device that write-back memory block act as for example, hard disk of the no disk client 21 to a certain extent.

For example, the memory headroom of the memory modules 32 is 8G, and based on some configuration strategies, the memory headroom for 8G It is configured as example, 4G memory field and 4G write-back memory block.Certainly, in different situations, the memory modules 32 memory headroom can be configured flexibly according to the situation of reality, for example the memory headroom for 8G is configured as example, 2G Memory field and 6G write-back memory block；Or the memory headroom for 8G be configured as example, 6G memory field and 2G write-back memory block etc., the not exhaustion herein of all so examples.

It should be strongly noted that different based on the hardware configuration without disk client 21, the internal memory of its memory modules 32 is empty Between may be 2G, 4G, 16G or 32G or even more, in the following description, in order to concisely illustrate the thought of the application And principle, temporarily illustrated so that the memory headroom of the memory modules 32 without disk client 21 is 8G as an example.It is in addition, described below Data either the size in space and capacity are generally described with modes such as 1G or 2G, what those skilled in the art could be aware that, Described 1G refers to that 1024M, described 2G refer to 2048M, and such it will not go into details.

The spatial information includes write-back memory block spatial information, memory field spatial information, write-back memory block Remaining space information, memory field remaining space information, write back request frequency, or current write-back amount.In embodiment, institute The write-back memory block spatial information and memory field spatial information stated are, for example, that 8G memory headroom is configured as example, 4G Memory field and 4G write-back memory block.The remaining space information of write data memory block or memory field is, for example, residue The information such as insufficient space 1024M or 512M；The write back request frequency is, for example, write-back in one minute or half a minute The frequency be more than 3 times or 2 times, such as when presently written data volume is 256M.But it is not limited thereto for description, in difference Application scenarios and example in can according to reality situation installation space, the frequency, the threshold value of data volume, will not be described here.

At least one without disk client described in the spatial information distribution that the distribute module 31 monitors according to the monitoring modular 30 The space-consuming of memory field and write-back memory block in the memory modules 32 at end 21.In the embodiment of reality, described point It is, for example, central processing unit without disk service end 11 or by one performed by the central processing unit or more with module 31 Combination of individual program or hardware and software etc..

The distribute module 31, which distributes the space of memory field and write-back memory block in the memory modules 32, includes two Kind situation, a kind of is that the space in memory modules 32 is allocated first, and another kind is, based on the allocated memory modules The space of memory field and write-back memory block in 32 is reallocated, that is, is redistributed.

In one embodiment in the first scenario, the distribute module 31 is used for occupied according to the memory field Space is to distribute the space of memory field and write-back memory block in the memory modules 32.Such as when the no disk client 21 The memory headrooms of memory modules 32 be 8G, wherein, based on the monitoring of the no disk client 21 to being currently running, its internal memory The program of area's operation has occupied 2.5G space, to ensure the internally stored program that may increase operation, then for the internal memory Then need to consider that the memory field is given in the space that memory field is enough, such as distribution 5G space during area's allocation space, remaining 3G's Space is as write-back memory block.

In an embodiment in the latter case, the distribute module 31 is based in the allocated memory modules 32 Memory field and write-back memory block space carry out space allocation, that is, redistribute.Such as when the no disk client 21 The memory headrooms of memory modules 32 be 8G, it is assigned with 4G spaces as memory field, 4G skies in addition for 8G memory headroom Between be used as write-back memory block, based on the monitoring of the no disk client 21 to being currently running, its whole internal memory is write in discovery The program run in area has only taken up 1.5G space, and estimate the program that it runs in certain period of time do not have it is too big Change, then can carry out space allocation to the space of the write-back memory block, for example mark off again from the memory field 1G distributes to the write-back memory block, that is, the space in the memory modules 32 redistributed is divided into 3G memory field and 5G Write-back memory block.

In the implementation process of above-mentioned two situations, it can be based on empty to write-back memory block spatial information and memory field Between the threshold value of information set to realize dynamic and non-dynamic distribution, such as, such as the internal memory mould when the no disk client 21 The memory headroom of block 32 is 8G, and it is assigned with 4G spaces for 8G memory headroom and is used as back as memory field, 4G spaces in addition Write data storage area, when the distribution of the gross space of internal memory must not be arranged to be less than 3G when, even if based on the institute to being currently running The monitoring for stating no disk client 21 finds that writing the program run in its whole memory field has only taken up 1.5G space, and estimates it The program run in certain period of time does not have too big change, but also the space of the write-back memory block will not be entered The larger space allocation of row, for example 2G will not be marked off from the memory field and distributes to the write-back memory block, and then The memory field of space division in the memory modules 32 for ensuring to redistribute will not be less than 3G space.

Due to the application without disk client 21 during operation, the write-back that is distributed from the space of memory modules 32 What data storage area act as the no disk client 21 to a certain extent is, for example, the storage device of hard disk, so that making Obtain without write-back is write in the storage device of the no disk server by network, and then alleviate net to a certain extent Network and the pressure without disk server；Further, since it is to deposit to the data that write operation write-in is carried out without disk client 21 based on user In the write-back memory block, this partial data that this allows for writing to the write-back memory block is described for storage The write-back memory block of memory modules 32 can be directly accessed to this when performing read operation again without disk client 21 and reading A little data carry out read operation without by storage device of the network access without disk service end 11, this is also avoided to a certain extent again The problems such as bringing reading or writing speed slow because of network bandwidth limitation.

In some embodiments of the application, the distribute module 31 also with when the monitoring modular 30 monitor it is described extremely When write-back memory block remaining space is less than certain threshold value in few one memory modules 32, the write-back being stored in is passed through Network storage to described at least one without the memory module 33 of disk service end 11 with discharge the occupancy of the write-back memory block sky Between.Such as when the memory headroom of the memory modules 32 of the no disk client 21 is 8G, it is assigned with 4G for 8G memory headroom Space as memory field, 4G spaces in addition as write-back memory block, the default threshold value be, for example, 1024M or 512M, when the distribute module 31 also monitors the remaining space of the write-back memory block less than 1024M or 512M, Data to ensure subsequently to write can have enough spaces to be written into the write-back memory block, then need to have been written into The data of the write-back memory block are transferred in no disk server by network, and then discharge the write-back memory block Space-consuming, it is 4G again to make it, and then meets demand of the follow-up write-in data to memory space.This implementation need not Memory headroom is redistributed, i.e., in the remaining space deficiency of write-back memory block and the space of memory field need not be seized Resource, data conversion storage simply is carried out according to default rule to write-back memory block and empties the write-back memory block.

In some embodiments of the application, the distribute module 31 can dynamically adjust the storage in memory modules 32 Space, i.e., described distribute module 31 are used for according to described in the monitoring modular 30 in real time monitoring at least one without in disk client 21 The space that memory field or the write-back memory block are actually taken in storing module 32 is with memory modules described in dynamically distributes The space of memory field and write-back memory block in 32.

In some embodiments of the application, the distribute module 31 be additionally operable to when the monitoring modular 30 monitor it is described Redistributed when write-back memory block remaining space is less than certain threshold value at least one memory modules 32 without disk client 21 The write-back memory block is in the space-consuming in the memory modules 32.Such as the internal memory when the no disk client 21 The memory headroom of module 32 is 8G, and it is assigned with 4G spaces as memory field, 4G spaces conduct in addition for 8G memory headroom Write-back memory block, based on the monitoring of the no disk client 21 to being currently running, find in its write-back memory block Remaining space has been less than certain threshold value, for example the default threshold value is, for example, 1024M or 512M, when the distribute module 31 When also monitoring the remaining space of the write-back memory block less than 1024M or 512M, based on to described in being currently running Monitoring without disk client 21, the program for finding to run in its whole memory field have only taken up 1.5G space, and estimate its The program of operation does not have too big change in certain period of time, then the space of the write-back memory block can be carried out empty Between adjust, for example mark off 1G from the memory field again and distribute to the write-back memory block, that is, the internal memory redistributed Space in module 32 is divided into 3G memory field and 5G write-back memory block.

In some embodiments of the application, the distribute module 31 be additionally operable to when the monitoring modular 30 monitor it is described extremely Lack one without the write back request frequency of disk client 21 to redistribute the occupancy of the write-back memory block in its memory modules 32 Space.Such as when the memory headroom of the memory modules 32 of the no disk client 21 is 8G, it is assigned with for 8G memory headroom 4G spaces are as memory field, and 4G spaces in addition are as write-back memory block, based on the no disk client to being currently running The monitoring at end 21, monitor that the no disk client 21 continually writes data to the write-back memory block, it is described frequent Refer to the frequency of write-back in such as one minute or half a minute more than 3 times or 2 times, then system is thought due to frequently writing behaviour The spatial content deficiency of write-back memory block in the memory modules 32 is may result in, and then needs to redistribute described Space-consuming of the write-back memory block in the memory modules 32, for example mark off 1G from the memory field again and distribute to The write-back memory block, that is, the space in the memory modules 32 redistributed are divided into 3G memory field and 5G write-back number According to memory block.Description for the write operation frequency is not limited thereto, can be according to reality in different application scenarios and example The situation on border sets the threshold value of the frequency, will not be described here.

In some embodiments of the application, the distribute module 31 be additionally operable to when the monitoring modular 30 monitor it is described extremely Lack a current write-back amount without disk client 21 to redistribute accounting for for the write-back memory block in its memory modules 32 Use space.

The current write-back amount such as be current time single write-back amount, such as the data volume that single writes is During 256M, then system think due to single write-in data volume it is larger, may subsequently still suffer from larger data volume be written into And the spatial content deficiency of write-back memory block in the memory modules 32 can be caused, and then need to redistribute the write-back Space-consuming of the data storage area in the memory modules 32.

The current write-back amount is such as the write-back quantity in current time or recent write-back amount, that is, is worked as Write-back amount in the previous period, when write-back amount is 256M in for example, current one minute, then system thinks Because lasting write-in data volume is larger, it may subsequently still suffer from larger data volume and be written into and then the internal memory mould can be caused The spatial content deficiency of write-back memory block in block 32, and then need to redistribute the write-back memory block described interior Space-consuming in storing module 32.

The current write-back amount is such as continuous write-back amount several times, write-back for example, twice in succession Measure for 256M when, then system thinks that data volume is larger due to being continuously written into, and may subsequently still suffer from larger data volume and be write Enter and then the spatial content deficiency of write-back memory block in the memory modules 32 can be caused, and then need to redistribute described Space-consuming of the write-back memory block in the memory modules 32.

Such as when the memory headroom of the memory modules 32 of the no disk client 21 is 8G, it is distributed for 8G memory headroom 4G spaces are as memory field, and 4G spaces in addition are as write-back memory block, based on the no disk visitor to being currently running The monitoring at family end 21, when monitoring that the current write-back amount of the no disk client 21 is more than a certain threshold value, this threshold value example Such as be 256M, then system think due to write-in data volume it is larger, may subsequently still suffer from larger data volume be written into and then The spatial content deficiency of write-back memory block in the memory modules 32 can be caused, and then need to redistribute the write-back number According to space-consuming of the memory block in the memory modules 32, for example mark off 1G from the memory field again and distribute to described return Data storage area is write, that is, the space in the memory modules 32 redistributed is divided into 3G memory field and 5G write-back storage Area.The description of data volume size for write operation write-in is not limited thereto, can be with different application scenarios and example The threshold value of writing, such as 128M, 64M etc. are set according to actual situation, will not be described here.

In some embodiments of the application, the distribute module 31 be additionally operable to when the monitoring modular 30 monitor it is described The internal memory is redistributed when memory field remaining space is less than certain threshold value at least one memory modules 32 without disk client 21 Space-consuming of the area in the memory modules 32.Such as when the memory headroom of the memory modules 32 of the no disk client 21 is 8G, it is assigned with 4G spaces for 8G memory headroom and is based on as memory field, 4G spaces in addition as write-back memory block To the monitoring for the no disk client 21 being currently running, monitor that memory field remaining space is less than 1G in the memory modules 32 When, to ensure the program that may subsequently run to the further demand of memory headroom, it is necessary to extend the space of the memory field, then Need to redistribute space-consuming of the write-back memory block in the memory modules 32, such as again from the write-back number The memory field is distributed to according to 2G is marked off in memory block, that is, the space in the memory modules 32 redistributed is divided into the interior of 6G Deposit area and 2G write-back memory block.Description for memory field remaining space deficiency situation is not limited thereto, in difference Application scenarios and example in can according to reality situation set remaining space threshold value, will not be described here.

In some embodiments of the application, the distribute module 31 is additionally operable to when the monitoring modular 30 monitors institute State memory field remaining space at least one memory modules 32 without disk client 21 and be less than the write-back that will be stored in during certain threshold value Data, into described at least one memory module 33 without disk service end 11, and are redistributed the memory field and existed by network storage Space-consuming in the memory modules 32.Such as when the memory headroom of the memory modules 32 of the no disk client 21 is 8G, It is assigned with 4G spaces as memory field for 8G memory headroom, other 4G spaces as write-back memory block, based on pair The monitoring for the no disk client 21 being currently running, monitor that memory field remaining space is less than 1G in the memory modules 32 When, to ensure the program that may subsequently run to the further demand of memory headroom, it is necessary to extend the space of the memory field, then Need to redistribute space-consuming of the write-back memory block in the memory modules 32, if the now write-back Because write-in data cause it not have the space of redundancy to distribute to the memory field in memory block, then need the write-back number According to the data stored in memory block by network storage to the no disk service end 11, in other words, in write-back memory block Data carry out unloading and empty the write-back memory block, so as to be it from new allocation space, such as again from described 2G is marked off in write-back memory block and distributes to the memory field, that is, the space in the memory modules 32 redistributed is divided into 6G memory field and 2G write-back memory block.Description for memory field remaining space deficiency situation is not limited thereto, The threshold value of remaining space can be set in different application scenarios and example according to the situation of reality, will not be described here.

The internal storage management system of the application can dynamically adjust memory field and write-back memory block in memory modules 32 Space so that without disk client 21 in operation according to internal memory actual occupancy situation, in real time dynamic adjustment write-back The space of memory block, and then ensure that no disk client 21 will not bring internal memory not because its memory modules 32 is taken by write-back Foot, while also meet and be locally stored and reading the demand of write-back without disk client 21, certainly, also cause without that will return Write data to write by network in the storage device of the no disk server, alleviate network to a certain extent and without disk server Pressure.

The application also provides a kind of EMS memory management process, is taken applied to by least one without disk client and at least one without disk In end network consisting system of being engaged in, wherein, described at least one is configured with memory modules without disk client, and described at least one without disk service End is configured with memory module.The communication network of the network system can be internet, one or more in-house networks, LAN (LAN), wide area network (WLAN), storage area network (SAN) etc., or its is appropriately combined.Again referring to Fig. 1, being shown as including nothing Disk service end 11 manages and the network system architecture schematic diagram without disk client 21, one as depicted without disk service end 11 by Multiple no disk clients 21 are managed by above-mentioned communication network simultaneously, the no disk client 21 is also referred to as no disk work Stand, the no disk client 21 is connected by network with without disk service end 11, and no disk client 21 is downloaded from without disk service end 11 Bootstrap, configuration file, kernel and image file etc., with the start-up operation system without disk client 21 of realization.It is such as described When operating system is (SuSE) Linux OS, wherein, the bootstrap be (SuSE) Linux OS in boot loader, institute Kernel is stated as the kernel in (SuSE) Linux OS, the image file is comprising init files in (SuSE) Linux OS Initrd compressed files, the configuration file include kernel title, kernel particular location, the title of image file, The particular location of image file, and start kernel parameter (for example, IP arrange parameters, start-up mode parameter, hot plug are set Standby parameter).

Referring to Fig. 5, Fig. 5 is shown as the flow signal of the EMS memory management process of the application in one embodiment, as schemed institute Show, the EMS memory management process comprises the following steps：

In the step s 100, the memory field and write-back number configured at least one memory modules without disk client described in monitoring According to the spatial information of memory block；In embodiment, the memory headroom of the memory modules is configured as memory field and write-back Memory block；The memory field is used to store from the bootstrap without disk service end down operation system, kernel, image file and match somebody with somebody File etc. is put, also for storage in corresponding program and file run without disk client needed for some softwares etc..The write-back number It is used to store based on user to the data without the progress write operation write-in of disk client according to memory block, in other words, without disk client During operation, what the write-back memory block act as the no disk client to a certain extent is, for example, depositing for hard disk Store up equipment.

For example, the memory headroom of the memory modules is 8G, and based on some configuration strategies, the memory headroom quilt for 8G It is configured to for example, 4G memory field and 4G write-back memory block.Certainly, in different situations, the memory modules Memory headroom can be configured flexibly according to the situation of reality, for example the memory headroom for 8G is configured as the interior of for example, 2G Deposit area and 6G write-back memory block；Or the memory headroom for 8G is configured as example, 6G memory field and 2G Write-back memory block etc., the not exhaustion herein of all so examples.It should be strongly noted that based on the hardware without disk client Configuration is different, and the memory headroom of its memory modules may be 2G, 4G, 16G or 32G or even more, in the following description, in order to The thought and principle of the application concisely is illustrated, is temporarily entered so that the memory headroom of the memory modules without disk client is 8G as an example Row explanation.In addition, either the size in space and capacity are generally described data described below with modes such as 1G or 2G, ability Field technique personnel could be aware that described 1G refers to that 1024M, described 2G refer to 2048M, and such it will not go into details.

In certain embodiments, the spatial information includes write-back memory block spatial information, memory field spatial information, Write-back memory block remaining space information, memory field remaining space information, write back request frequency, or current write-back Amount.In embodiment, described write-back memory block spatial information and memory field spatial information are, for example, 8G memory headroom It is configured as example, 4G memory field and 4G write-back memory block.In embodiment, write data memory block or interior The remaining space information for depositing area is, for example, remaining space less than information such as 1024M or 512M；The write back request frequency is for example It is more than 3 times or 2 times for the frequency of write-back in one minute or half a minute, such as when presently written data volume is 256M. But it is not limited thereto for description, can be according to actual situation installation space, frequency in different application scenarios and example Secondary, data volume threshold value, will not be described here.

In step S101, according to internal memory at least one memory modules without disk client described in spatial information distribution Area and the space-consuming of write-back memory block.

In some different embodiments, the space of memory field and write-back memory block in the memory modules is distributed Including two kinds of situations, a kind of is that the space in memory modules is allocated first, and another kind is, based on the allocated internal memory The space of memory field and write-back memory block in module is reallocated, that is, is redistributed.

In one embodiment in the first scenario, according to the occupied space in the memory field to distribute the internal memory The space of memory field and write-back memory block in module.Such as when the memory headroom of the memory modules of the no disk client is 8G, wherein, based on the monitoring of the no disk client to being currently running, the program of its memory field operation has occupied 2.5G Space, the internally stored program of operation may be increased to ensure, then then to need to consider internal memory during the memory field allocation space The memory field is given in the enough spaces in area, such as distribution 5G space, and remaining 3G space is as write-back memory block.

In an embodiment in the latter case, based on the memory field in the allocated memory modules and write-back number Space allocation is carried out according to the space of memory block, that is, is redistributed.Such as when the internal memory of the memory modules of the no disk client is empty Between be 8G, it is assigned with 4G spaces as memory field for 8G memory headroom, other 4G spaces as write-back memory block, Based on the monitoring of the no disk client to being currently running, it is found that writing the program run in its whole memory field only takes up 1.5G space, and estimate the program that it runs in certain period of time and do not have too big change, then can be to the write-back The space of data storage area carries out space allocation, for example from the memory field marks off 1G again and distribute to the write-back and deposit Storage area, that is, the space in the memory modules redistributed are divided into 3G memory field and 5G write-back memory block.

In the implementation process of above-mentioned two situations, it can be based on empty to write-back memory block spatial information and memory field Between the threshold value of information set to realize dynamic and non-dynamic distribution, such as, such as memory modules when the no disk client Memory headroom be 8G, it is assigned with 4G spaces as memory field for 8G memory headroom, and 4G spaces in addition are as write-back number According to memory block, when the distribution of the gross space of internal memory must not be arranged to be less than 3G when, even if based on the nothing to being currently running The monitoring of disk client 21 finds that writing the program run in its whole memory field has only taken up 1.5G space, and estimates it one The program of operation does not have too big change in section of fixing time, but the space of the write-back memory block will not also be carried out compared with Big space allocation, for example 2G will not be marked off from the memory field and distributes to the write-back memory block, and then ensure The memory field of space division in the memory modules redistributed will not be less than 3G space.

Due to the application without disk client during operation, the write-back distributed from the space of memory modules What memory block act as the no disk client to a certain extent is, for example, the storage device of hard disk, since then so that without Write-back is write in the storage device of the no disk server by network, and then alleviates network and nothing to a certain extent The pressure of disk server；Further, since the data that write operation write-in is carried out without disk client are stored in based on user described In write-back memory block, this allows for write-in to this partial data of the write-back memory block by the no disk client End can directly access memory modules write-back memory block when performing read operation again and reading is read by those data Operate without again by storage device of the network access without disk service end, this is also avoided because Netowrk tape is extended the deadline to a certain extent The problems such as system brings reading or writing speed slow.

In certain embodiments, when write-back memory block at least one memory modules without disk client described in monitoring When remaining space is less than certain threshold value, by the write-back being stored in by network storage to described at least one without disk service end Memory module is to discharge the space-consuming of the write-back memory block.Such as when in the memory modules of the no disk client It is 8G to deposit space, and it is assigned with 4G spaces for 8G memory headroom and deposited as memory field, 4G spaces in addition as write-back Storage area, the default threshold value are, for example, 1024M or 512M, when monitoring the remaining space of the write-back memory block not During sufficient 1024M or 512M, the data to ensure subsequently to write can have enough spaces to be written into the write-back memory block, Then need the data for having been written into the write-back memory block being transferred in no disk server by network, and then discharge institute The space-consuming of write-back memory block is stated, it is 4G again to make it, and then meets need of the follow-up write-in data to memory space Ask.This implementation is unnecessary to redistribute memory headroom, i.e., in the remaining space deficiency of write-back memory block not Need to seize the space resources of memory field, data conversion storage simply is carried out according to default rule to write-back memory block and emptied The write-back memory block.

In some embodiments of the application, the storage that the present processes can also be dynamically adjusted in memory modules is empty Between, i.e., according to memory field or the write-back memory block at least one memory modules without disk client described in monitoring in real time The space actually taken is with the memory field in memory modules described in dynamically distributes and the space of write-back memory block.Refer to Fig. 6, the schematic flow sheet of the EMS memory management process of the application in another embodiment is shown as, as illustrated, the internal memory pipe Reason method comprises the following steps：

In step S300, memory field or described time at least one memory modules without disk client described in monitoring in real time Write the space that data storage area is actually taken., can be by the distribute module without configuration inside disk client in an embodiment The space that memory field or the write-back memory block are actually taken in its memory modules is monitored in real time；In another embodiment In, internal memory in the memory modules without disk client can be monitored in real time by the monitoring modular configured without disk service end by it The area space that either the write-back memory block is actually taken monitors multiple no disk clients each internal memory mould simultaneously The space that memory field or the write-back memory block are actually taken in block.

In step S301, according to described in the spatial information dynamically distributes monitored in real time at least one without disk client The space-consuming of memory field and write-back memory block in memory modules., can be by inside without disk client in an embodiment The distribute module of configuration is according to memory field and write-back in described its memory modules of spatial information dynamically distributes monitored in real time The space-consuming of memory block；, can be real-time by the distribute module foundation configured without disk service end by it in another embodiment Memory field and write-back memory block in one memory modules without disk client of the spatial information dynamically distributes of monitoring Space-consuming, or simultaneously to the space of memory field and write-back memory block in multiple no respective memory modules of disk client Carry out dynamically distributes.

In some embodiments of the application, described in the spatial information dynamically distributes that the foundation monitors in real time at least It is the step of the space-consuming of memory field and write-back memory block in one memory modules without disk client：Described in monitoring extremely Described return is redistributed when write-back memory block remaining space is less than certain threshold value in few memory modules without disk client Data storage area is write in the space-consuming in the memory modules.Such as the internal memory of the memory headroom when the no disk client Space is 8G, and it is assigned with 4G spaces for 8G memory headroom and stored as memory field, 4G spaces in addition as write-back Area, based on the monitoring of the no disk client to being currently running, it is found that remaining space is less than in its write-back memory block Certain threshold value, such as the default threshold value are, for example, 1024M or 512M, when the distribute module also monitors the write-back When the remaining space of data storage area is less than 1024M or 512M, based on the monitoring of the no disk client to being currently running, It was found that the program run in its whole memory field has only taken up 1.5G space, and estimate its journey run in certain period of time Sequence does not have too big change, then can carry out space allocation to the space of the write-back memory block, such as again from described 1G is configured in memory field and distributes to the write-back memory block, that is, the space configuration in the memory headroom redistributed is 3G Memory field and 5G write-back memory block.

At least one memory modules without disk client described in the spatial information dynamically distributes that the foundation monitors in real time The step of space-consuming of middle memory field and write-back memory block is：At least one write back request without disk client described in monitoring Frequency is to redistribute the space-consuming of the write-back memory block in its memory modules.Such as when in the no disk client The memory headroom for depositing space is 8G, and it is assigned with 4G spaces as memory field, 4G spaces conduct in addition for 8G memory headroom Write-back memory block, based on the monitoring of the no disk client to being currently running, monitor that the no disk client is frequent Ground writes data, the frequency for frequently referring to write-back in such as one minute or half a minute to the write-back memory block Secondary to be more than 3 times or 2 times, then system thinks to store because frequently write operation may result in write-back in the memory headroom The spatial content deficiency in area, and then need to redistribute occupancy sky of the write-back memory block in the memory headroom Between, for example configure 1G from the memory field again and distribute to the write-back memory block, that is, the memory headroom redistributed In space configuration be 3G memory field and 5G write-back memory block.Description for the write operation frequency is not limited to This, can set the threshold value of the frequency according to the situation of reality in different application scenarios and example, will not be described here.

In some embodiments of the application, described in the spatial information dynamically distributes that the foundation monitors in real time at least It is the step of the space-consuming of memory field and write-back memory block in one memory modules without disk client：Described in monitoring at least The one current write-back amount without disk client is to redistribute the space-consuming of the write-back memory block in its memory modules.

The current write-back amount such as be current time single write-back amount, such as the data volume that single writes is During 256M, then system think due to single write-in data volume it is larger, may subsequently still suffer from larger data volume be written into And the spatial content deficiency of write-back memory block in the memory modules can be caused, and then need to redistribute the write-back number According to space-consuming of the memory block in the memory modules.

The current write-back amount is such as the write-back quantity in current time or recent write-back amount, that is, is worked as Write-back amount in the previous period, when write-back amount is 256M in for example, current one minute, then system thinks Because lasting write-in data volume is larger, it may subsequently still suffer from larger data volume and be written into and then the internal memory mould can be caused The spatial content deficiency of write-back memory block in block, and then need to redistribute the write-back memory block in the internal memory Space-consuming in module.

The current write-back amount is such as continuous write-back amount several times, write-back for example, twice in succession Measure for 256M when, then system thinks that data volume is larger due to being continuously written into, and may subsequently still suffer from larger data volume and be write Enter and then the spatial content deficiency of write-back memory block in the memory modules can be caused, and then need to redistribute described return Write space-consuming of the data storage area in the memory modules.Such as when the internal memory of the memory headroom of the no disk client is empty Between be 8G, it is assigned with 4G spaces as memory field for 8G memory headroom, other 4G spaces as write-back memory block, Based on the monitoring of the no disk client to being currently running, monitor that the current write-back amount of the no disk client is more than During a certain threshold value, this threshold value is, for example, 256M, then system think due to write-in data volume it is larger, may subsequently still suffer from compared with Big data volume is written into and then can cause the spatial content deficiency of write-back memory block in the memory headroom, and then needs Redistribute space-consuming of the write-back memory block in the memory headroom, for example configure again from the memory field Go out 1G and distribute to the write-back memory block, that is, the space configuration in the memory headroom redistributed is 3G memory field and 5G Write-back memory block.Description for the data volume size of write operation write-in is not limited thereto, in different applied fields The threshold value of writing, such as 128M, 64M etc. can be set in scape and example according to the situation of reality, will not be described here.

In some embodiments of the application, described in the spatial information dynamically distributes that the foundation monitors in real time at least In one memory modules without disk client the step of the space-consuming of memory field and write-back memory block for monitor it is described extremely The memory field is redistributed in institute when memory field remaining space is less than certain threshold value in few memory modules without disk client State the space-consuming in memory modules.Such as when the memory headroom of the memory headroom of the no disk client is 8G, it is 8G's Memory headroom is assigned with 4G spaces as memory field, and 4G spaces in addition are as write-back memory block, based on to being currently running The no disk client monitoring, it is follow-up to ensure when monitoring that memory field remaining space is less than 1G in the memory headroom May operation program to the further demand of memory headroom, it is necessary to extend the space of the memory field, then need to redistribute Space-consuming of the write-back memory block in the memory headroom, for example configure again from the write-back memory block Go out 2G and distribute to the memory field, that is, the space configuration in the memory headroom redistributed is 6G memory field and 2G write-back number According to memory block.Description for memory field remaining space deficiency situation is not limited thereto, in different application scenarios and example In can according to reality situation set remaining space threshold value, will not be described here.

In some embodiments of the application, described in the spatial information dynamically distributes that the foundation monitors in real time at least It is described to monitor the step of the space-consuming of memory field and write-back memory block in one memory modules without disk client The write-back being stored in is led to when memory field remaining space is less than certain threshold value at least one memory modules without disk client Network storage is crossed at least one memory module without disk service end, and redistributes the memory field in the internal memory mould Space-consuming in block.Such as when the memory headroom of the memory headroom of the no disk client is 8G, it is 8G memory headroom 4G spaces are assigned with as memory field, 4G spaces in addition are as write-back memory block, based on the nothing to being currently running The monitoring of disk client, subsequently may operation to ensure when monitoring that memory field remaining space is less than 1G in the memory headroom Program to the further demand of memory headroom, it is necessary to extend the space of the memory field, then need to redistribute the write-back Space-consuming of the data storage area in the memory headroom, if now because write-in data cause in the write-back memory block It does not have the space of redundancy to distribute to the memory field, then needs to lead to the data stored in the write-back memory block Network storage is crossed to the no disk service end, in other words, the data in write-back memory block is carried out with unloading and is emptied described Write-back memory block, so as to be it from new allocation space, for example from the write-back memory block configure 2G again The memory field is distributed to, that is, the space configuration in the memory headroom redistributed is that 6G memory field and 2G write-back are deposited Storage area.Description for memory field remaining space deficiency situation is not limited thereto, can in different application scenarios and example To set the threshold value of remaining space according to actual situation, will not be described here.

The EMS memory management process of the application can dynamically adjust memory field and write-back memory block in memory modules Space so that the actual occupancy situation according to internal memory, the write-back of dynamic adjustment in real time store in operation without disk client The space in area, and then ensure that no disk client will not bring low memory because its memory modules is taken by write-back, simultaneously Also meet and be locally stored and reading the demand of write-back without disk client, certainly, also cause without write-back is led to Cross network to write in the storage device of the no disk server, alleviate network and the pressure without disk server to a certain extent.

The application also provides a kind of computer-readable recording medium, is stored thereon with memory management computer program, described When memory management computer program is executed by processor the step of above-mentioned EMS memory management process.

In the application provide embodiment in, it is described it is computer-readable write storage medium can include RAM, ROM, EEPROM, CD-ROM or other optical disk storage apparatus, disk storage device or other magnetic storage apparatus, flash memory or it can use Any other of line access can be entered in desired program code of the storage with instruction or data structure form and by computer Medium.In addition, any connection can be properly termed as computer-readable medium.If for example, instruction be using coaxial cable, The wireless technology of optical fiber cable, twisted-pair feeder, digital subscriber line (DSL) or such as infrared ray, radio and microwave etc, from net Stand, server or other remote sources send, then the coaxial cable, optical fiber cable, twisted-pair feeder, DSL or such as infrared ray, The wireless technology of radio and microwave etc is included in the definition of the medium.It is it is to be understood, however, that computer-readable Writing storage medium and data storage medium does not include connection, carrier wave, signal or other fugitive mediums, and is intended to be directed to Non-transitory, tangible storage medium.As the disk used in application and CD include compact disk (CD), laser-optical disk, CD, digital versatile disc (DVD), floppy disk and Blu-ray Disc, wherein, disk generally magnetically replicate data, and CD is then With laser come optically replicate data.

In one or more illustrative aspects, the function described by herein described memory management computer program can be used Hardware, software, mode of firmware or its any combination are realized.When implemented in software, can be using these functions as one Individual or multiple instruction or code storage are sent on computer-readable medium.The step of method disclosed in the present application or algorithm Software module can be can perform with processor to embody, wherein the executable software module of processor can be located at tangible, non-provisional Property is computer-readable to be write in storage medium.Tangible, the computer-readable storage medium of writing of non-transitory can be that computer can be deposited Any usable medium taken.

The application can be described in the general context of computer executable instructions, such as program Module.Usually, program module includes performing particular task or realizes routine, program, object, the group of particular abstract data type Part, data structure etc..The application can also be put into practice in a distributed computing environment, in these DCEs, by Task is performed and connected remote processing devices by communication network.In a distributed computing environment, program module can be with In the local and remote computer-readable storage medium including storage device.

In summary, due to the application provide embodiment without disk client during operation, from memory modules Space in the write-back memory block distributed act as the no disk client to a certain extent be, for example, hard disk storage Equipment, since then so that without write-back is write in the storage device of the no disk server by network, Jin Eryi Determine to alleviate network and the pressure without disk server in degree；Further, since based on user to carrying out write operation without disk client The data of write-in are stored in the write-back memory block, this allow for write-in to the write-back memory block this Partial data performs read operation again by the no disk client and the write-back of memory modules can be directly accessed when reading Memory block carries out read operation to those data without again by storage device of the network access without disk service end, this also certain journey Avoided on degree because network bandwidth limits the problems such as bringing reading or writing speed slow.

Memory field and write-back in memory modules can be dynamically adjusted and deposit by being additionally, since the application and providing embodiment The space of storage area so that without disk client in operation according to internal memory actual occupancy situation, in real time dynamic adjustment write-back number According to the space of memory block, and then ensure that no disk client will not bring internal memory not because its memory modules is taken by write-back Foot, while also meet and be locally stored and reading the demand of write-back without disk client, certainly, also causing need not be by write-back Data are write by network in the storage device of the no disk server, alleviate network to a certain extent and without disk server Pressure.

The principle and its effect of above-described embodiment only illustrative the application, not for limitation the application.It is any ripe Know the personage of this technology all can without prejudice to spirit herein and under the scope of, modifications and changes are carried out to above-described embodiment.Cause This, those of ordinary skill in the art is complete without departing from spirit disclosed herein and institute under technological thought such as Into all equivalent modifications or change, should be covered by claims hereof.

Claims

1. a kind of without disk client, applied to at least one network system be connected without disk service end, it is characterised in that wrap Include：

Memory modules, its memory headroom are configured as memory field and write-back memory block；

Execution module, it is configurable for reading data or write-back without disk service end to described at least one by network；With And it is configured as the data to be written storage when receiving write back request and arrives the write-back memory block；

Distribute module, for distributing the memory modules according to the occupied space in the write-back memory block or memory field Middle memory field and the space of write-back memory block.

2. no disk client according to claim 1, it is characterised in that the distribute module is additionally operable to monitor in described When write-back memory block remaining space is less than certain threshold value in storing module by the write-back being stored in by network storage extremely The no disk service end is to discharge the space-consuming of the write-back memory block.

3. no disk client according to claim 1, it is characterised in that the distribute module is used to monitor in real time in described Deposit space that area or the write-back memory block actually taken with the memory field in memory modules described in dynamically distributes and The space of write-back memory block.

4. no disk client according to claim 3, it is characterised in that the distribute module is additionally operable to monitor in described The write-back memory block is redistributed when write-back memory block remaining space is less than certain threshold value in storing module described Space-consuming in memory modules.

5. no disk client according to claim 3, it is characterised in that the distribute module is additionally operable to monitor the execution Module write back request frequency or current write-back amount are to redistribute the write-back memory block in the memory modules In space-consuming.

6. no disk client according to claim 3, it is characterised in that the distribute module is additionally operable to monitor in described Occupancy of the memory field in the memory modules is redistributed when memory field remaining space is less than certain threshold value in storing module Space.

7. no disk client according to claim 3, it is characterised in that the distribute module is additionally operable to monitor in described The write-back being stored in is passed through into network storage to the no disk when memory field remaining space is less than certain threshold value in storing module Service end, and redistribute space-consuming of the memory field in the memory modules.

8. one kind is without disk service end, at least one no disk client of the management configuration in network system, it is characterised in that Including：

Memory module, the configuration processor and described at least one for being read by described at least one without disk client is stored with without disk visitor The write-back that family end is write by network transmission；

Distribute module, at least one memory headroom without disk client to be configured into memory field and write-back storage Area, memory field and write-back memory block are occupied according to the write-back memory block or memory field in the memory headroom Space be allocated.

9. no disk service end according to claim 8, it is characterised in that the distribute module be additionally operable to monitor it is described extremely Returned when write-back memory block remaining space is less than certain threshold value in a few memory headroom without disk client by what is be stored in Write data by network storage to the memory module to discharge the space-consuming of the write-back memory block.

10. no disk service end according to claim 8, it is characterised in that the distribute module is used for described in monitoring in real time The space that memory field or the write-back memory block are actually taken is with the memory field in memory headroom described in dynamically distributes And the space of write-back memory block.

11. no disk service end according to claim 10, it is characterised in that the distribute module is additionally operable to monitor described Institute is redistributed when write-back memory block remaining space is less than certain threshold value in the memory headroom of at least one no disk client Write-back memory block is stated in the space-consuming in the memory headroom.

12. no disk service end according to claim 10, it is characterised in that the distribute module is additionally operable to described in monitoring extremely Lack a write back request frequency without disk client or current write-back amount to redistribute the write-back number in its memory headroom According to the space-consuming of memory block.

13. no disk service end according to claim 10, it is characterised in that the distribute module is additionally operable to monitor described The memory field is redistributed when memory field remaining space is less than certain threshold value at least one memory headroom without disk client to exist Space-consuming in the memory headroom.

14. no disk service end according to claim 10, it is characterised in that the distribute module is additionally operable to monitor institute State memory field remaining space at least one memory headroom without disk client and be less than the write-back that will be stored in during certain threshold value By network storage to the memory module, and redistribute space-consuming of the memory field in the memory headroom.

15. a kind of internal storage management system, applied to by least one without disk client and at least one without disk service end network consisting In system, wherein, described at least one is configured with memory modules without disk client, and described at least one is configured with storage without disk service end Module, it is characterised in that including：

Monitoring modular, deposited for monitoring the memory field configured at least one memory modules without disk client and write-back The spatial information of storage area；

Distribute module, according at least one memory modules without disk client described in the spatial information distribution of the monitoring module monitors Middle memory field and the space-consuming of write-back memory block.

16. internal storage management system according to claim 15, it is characterised in that the spatial information is deposited including write-back Storage area spatial information, memory field spatial information, write-back memory block remaining space information, memory field remaining space information, return Write request frequency, or current write-back amount.

17. internal storage management system according to claim 15, it is characterised in that the distribute module, which is also used, works as the monitoring Module monitors write-back memory block remaining space at least one memory modules without disk client is less than certain threshold value When, by the write-back being stored in by network storage at least one memory module without disk service end to discharge described return Write the space-consuming of data storage area.

18. internal storage management system according to claim 15, it is characterised in that the distribute module is used for according to the prison Survey memory field or write-back memory block quilt at least one memory modules without disk client described in the real-time monitoring of module The space actually taken is with the memory field in memory modules described in dynamically distributes and the space of write-back memory block.

19. internal storage management system according to claim 18, it is characterised in that the distribute module is additionally operable to work as the prison Survey module monitors write-back memory block remaining space at least one memory modules without disk client and be less than certain threshold The write-back memory block is redistributed during value in the space-consuming in the memory modules.

20. internal storage management system according to claim 18, it is characterised in that the distribute module is additionally operable to work as the prison At least one write back request frequency without disk client described in module monitors or current write-back amount are surveyed to redistribute in it The space-consuming of write-back memory block in storing module.

21. internal storage management system according to claim 18, it is characterised in that the distribute module is additionally operable to work as the prison Survey when module monitors memory field remaining space at least one memory modules without disk client is less than certain threshold value again Distribute space-consuming of the memory field in the memory modules.

22. internal storage management system according to claim 18, it is characterised in that the distribute module is additionally operable to work as the prison Module monitors are surveyed to incite somebody to action when at least one memory modules without disk client, memory field remaining space is less than certain threshold value The write-back being stored in, at least one memory module without disk service end, and is redistributed described by network storage Space-consuming of the memory field in the memory modules.

23. a kind of EMS memory management process, applied to by least one without disk client and at least one without disk service end network consisting In system, wherein, described at least one is configured with memory modules without disk client, and described at least one is configured with storage without disk service end Module, it is characterised in that comprise the following steps：

The memory field and the space letter of write-back memory block configured at least one memory modules without disk client described in monitoring Breath；

According to memory field at least one memory modules without disk client described in spatial information distribution and write-back storage The space-consuming in area.

24. EMS memory management process according to claim 23, it is characterised in that the spatial information is deposited including write-back Storage area spatial information, memory field spatial information, write-back memory block remaining space information, memory field remaining space information, return Write request frequency, or current write-back amount.

25. EMS memory management process according to claim 23, it is characterised in that described to distribute institute according to the spatial information The step of stating the space-consuming of memory field and write-back memory block at least one memory modules without disk client be：Work as monitoring When at least one memory modules without disk client, write-back memory block remaining space is less than certain threshold value, it will deposit The write-back entered is deposited by network storage at least one memory module without disk service end with discharging the write-back The space-consuming of storage area.

26. EMS memory management process according to claim 23, it is characterised in that：

Memory field or the write-back memory block are by reality at least one memory modules without disk client described in monitoring in real time The space that border takes；

According to memory field at least one memory modules without disk client described in the spatial information dynamically distributes monitored in real time And the space-consuming of write-back memory block.

27. EMS memory management process according to claim 26, it is characterised in that the space that the foundation monitors in real time The space-consuming of memory field and write-back memory block at least one memory modules without disk client described in information dynamically distributes The step of be：Monitor that write-back memory block remaining space is less than certain at least one memory modules without disk client The write-back memory block is redistributed during threshold value in the space-consuming in the memory modules.

28. EMS memory management process according to claim 26, it is characterised in that the space that the foundation monitors in real time The space-consuming of memory field and write-back memory block at least one memory modules without disk client described in information dynamically distributes The step of be：At least one write back request frequency without disk client or current write-back amount described in monitoring are to redistribute it The space-consuming of write-back memory block in memory modules.

29. EMS memory management process according to claim 26, it is characterised in that the space that the foundation monitors in real time The space-consuming of memory field and write-back memory block at least one memory modules without disk client described in information dynamically distributes The step of be：Monitor weight when memory field remaining space is less than certain threshold value at least one memory modules without disk client Newly distribute space-consuming of the memory field in the memory modules.

30. EMS memory management process according to claim 26, it is characterised in that the space that the foundation monitors in real time The space-consuming of memory field and write-back memory block at least one memory modules without disk client described in information dynamically distributes The step of be：When monitoring that memory field remaining space is less than certain threshold value at least one memory modules without disk client By the write-back being stored in by network storage at least one memory module without disk service end, and redistribute institute State space-consuming of the memory field in the memory modules.

31. a kind of computer-readable recording medium, it is stored thereon with memory management computer program, it is characterised in that in described Deposit the step of realizing any one of the claim 23-30 EMS memory management process when management computer program is executed by processor.