CN103702057B - Block storage algorithm applicable to multiple paths of concurrent-written stream media data - Google Patents
Block storage algorithm applicable to multiple paths of concurrent-written stream media data Download PDFInfo
- Publication number
- CN103702057B CN103702057B CN201310394300.8A CN201310394300A CN103702057B CN 103702057 B CN103702057 B CN 103702057B CN 201310394300 A CN201310394300 A CN 201310394300A CN 103702057 B CN103702057 B CN 103702057B
- Authority
- CN
- China
- Prior art keywords
- medium data
- headend equipment
- stream medium
- block
- buffer area
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Television Signal Processing For Recording (AREA)
- Signal Processing For Digital Recording And Reproducing (AREA)
Abstract
The invention relates to a block storage algorithm applicable to multiple paths of concurrent-written stream media data. In the security and protection monitoring industry, requirements on the storage property of a video server are higher and higher as the project scale is continuously expanded and a high-definition front-end camera is used. A bottleneck formed by concurrent writing of hundreds of external video data of the camera into a single storage server is analyzed. A file system breaks down easily and the data is seriously lost under the abnormal condition of the server. After a disk is repeatedly covered and written, a large number of disk fragmentations are generated and the storage property is also reduced. According to the invention, the problems are analyzed, and a strategy for establishing a self-cache system through bypassing a file system layer, as well as a block storage idea and an available block searching algorithm on the basis of the block storage idea are provided. Through the online use of a practical product, the feasibility and the reliability of the block storage algorithm are verified.
Description
Technical field
Present invention is mainly applied to security protection Network Video Surveillance field, is specifically designed for the stream matchmaker produced in Network Video Surveillance
Volume data, there is provided a kind of storage algorithm of optimization, the algorithm can be applicable in Network Video Surveillance in all of storage device,
To ensure efficient, the optimization storage of data.
Background technology
In recent years, with building security protection, garden security protection and the construction of safe city, the video monitoring of networking is presented
Except rapid growth trend, from SD to high definition, the stream medium data that web camera is produced is passed video data by network
It is defeated to video server when, needs store and preserve certain number of days, to use when some events occur
Family playback video recording.Thus, it is stored in that proportion shared in network monitoring is also increasing, the safety and stability of its data is outstanding
For important, so as to the memory requirement of networked video also with regard to more and more higher.
Existing memory technology is all based on the storage mode of system default, and its basic storage thinking is：Institute
Some files are all built on the virtual file system VFS of operating system, when system has video data to write, first,
System will store into file little one by one the video data for transmitting, and the Production conditions of new file can be defined
（Such as when a hour or file size are more than 200MB）, so, when the conditions are met, Virtual File System VFS will
Produce a new file to carry out the storage of video data；Then, video data is write into into disk by way of file high
Speed caching（Disk Caches）, so as to by the data buffer storage on disk in internal memory；Finally, system is irregular by disk
Cache to hard disk and write data, stored.
Need for continual storage for a large amount of stream medium datas in Network Video Surveillance, this kind of storage scheme
There is following drawback.
Firstly, since before network video data storage, need to first pass through disk cache and stored, when front end is deposited
When a large amount of web cameras carry out data acquisition and need storage, certain video data was needed first by it before write hard disk
Corresponding caching is filled, and when the space of the buffer zone storage is finished, then could write hard disk, empties caching.By
Determined by system default in the caching distributed for each network video stream, thus the hard disk write time of caching can not be true
It is fixed, so, once when the phenomenons such as the machine of delaying, power down occurs in server, the data in disk cache will be unable to write hard disk, delay
Deposit data is all lost, it is most likely that a large amount of disappearances of certain segment data can be caused, so as to bring serious loss.
Secondly, when system cache writes video data into hard disk, because system adopts the piecemeal storage mode of acquiescence, by
It is uncontrolled in the size of each file, when file size was specified too as a child, due to the data volume of the generation of high-definition camera
It is larger, such as 4Mbps, so, video data will be dispersed into many small documents in hard-disk system and carry out distribution storage, so,
Due to frequently opening and closing file, so as to cause extra overhead, cause effectiveness of retrieval to reduce, affect system
Operating experience；If conversely, regulation file size is too big, because system is that to be responsible for finding available disk by operating system empty
Between, file distribution multiple different positions on disk will be so caused, so that the disk tracking time increases, disk read-write
Speed is reduced, and affects systematic function.
3rd, because disk cache to hard disk when data are write, its write cycle is according to each spatial cache
It is different and different, and each spatial cache is dynamic change, so as to, DISK to Image group is nearly always at driving condition,
Power consumption is caused to increase, hard disk life-span is reduced.
4th, in safety monitoring industry, the continuous expansion of scale of the project, the use of high definition front-end camera, front end regards
Frequency code stream is continuously increased, and system default storage mode occurs bottleneck, causes its video code flow for being at best able to process not surpass
150Mbps is crossed, so for separate unit storage device, outside there are hundreds of road even thousands of road camera video data to need concurrently to write
When entering server, then the quantity for increasing server is needed to complete storage demand.
Finally, due to the particularity of security protection industry, needs the permanently effective operation of monitoring device, thus, the stability of equipment
It is particularly important, and the particularity of video monitoring data, it is that it will constantly carry out substantial amounts of storage and deleted issue
According to, thus, in traditional storage mode, when system cache writes video data into hard disk, because system is using dividing for giving tacit consent to
Block storage mode, so, is deleted the hard drive space of data, will not be able to be utilized again during use afterwards,
Cause substantial amounts of disk fragmentses to produce, have a strong impact on systematic function, on the one hand, obstacle occurs in the write that can cause data, causes
Loss of data；On the other hand, fragment can excessively affect whole file system, cause the hydraulic performance decline of whole system, it is necessary to close
System, carries out Disk Cleanup, and moreover, substantial amounts of fragment can cause hard disk utilization to reduce（Jing is counted, and hard disk utilization is most
More not over 80%）Reduce with hard disk life-span, this will cause the cost for storing to raise.
The content of the invention
The purpose of the present invention is that the network video data storage solved the problems, such as in whole security protection industry, and the present invention will
Overcome the shortcoming in above-mentioned memory technology such that it is able to provide a kind of side for being suitable for network video data storage in security protection industry
Method, the wisdom storage to reach efficient, stable, energy-conservation, without fragment.
To achieve these goals, the present invention proposes a kind of sectional type for being adapted to multipath concurrence write stream medium data
The thought of algorithm is stored, its basic thought is：In security protection industry, network video data is otherwise known as stream medium data, works as stream
When media data is in storing process, write data bandwidth is much larger than reading data bandwidth, that is to say, that system overwhelming majority band
Wide expense seldom has the data to being most recently written to be read out application all in data receiver storage, so the present invention gets rid of
File system storage mode is abandoned, disk has been opened in direct mode, got around disk cache, the region that directly disk is specified
It is written and read operation.In other words, this algorithm has invented one kind by caching system（Cache System）It is more conducive to using a set of
The special caching mechanism that stream medium data is used, to the disk general caching mechanism for replacing kernel to provide.This storage algorithm
Realize that step is described as follows.
First, caching system is initialized, caching system is distributed in most of space of Installed System Memory, according to buddy system
Algorithm is managed to storage allocation, slow when being each headend equipment distribution write memory block with reference to the characteristics of stream medium data
Deposit.One important indicator of stream medium data writing speed is exactly code stream, and it illustrates headend equipment generation data volume per second
Size, and headend equipment code stream and unstable, it is therefore desirable to calculate the video flow of headend equipment generation per second（Code stream）The number of Q
Term hopes.Distribute C times of the constant that cache size Fb arranges code stream for headend equipment, its expression system can buffer the data of C seconds,
Formula is as follows：
Second, hard disk storage system is initialized, disk is opened in direct mode, by disk according to fixed size, respectively
Into several memory blocks（chunk）, each headend equipment to the specified memory block distributed for it writes data, when memory block writes full
Afterwards according to available block probe algorithm（Give an account of as follows and continue）Search available block.System is located in the way of the multi-path asynchronous write of single thread
The problem of reason multipath concurrence write, effectively reduces multithreading expense so that system is more stable.
Available block probe algorithm is described.
When headend equipment to write full for its memory block distributed after, search available block according to available block probe algorithm.Assume
Memory block number total in system is T, and it is E (S) that all headend equipments produce the mathematic expectaion of total code stream, current headend equipment code
Proportion of the stream in total code stream is P=upper (E (S)/E (Q)), has write full memory block a, is detecting next available block,
Probe algorithm Find Chunk () is described as follows：
Step 1：Initializing variable cyclic variable i=0, k=1, temporary variable nTmpBlock=0；
Step 2：Whether cyclic variable i value is judged less than P, if condition is set up to continue to perform downwards；
Step 3：Seek next possible available memory block nActBlock：nTmpBlock = a + (P – i ) * k；
NActBlock=nTmpBlock % total block data T；
Step 4：Whether nTmpBlock is judged less than current block a+ total block data T, if condition is false jumps to step 6, if bar
Part is set up and continues to perform downwards；
Step 5：Judge that whether nActBlock is not used by remaining headend equipment, or the video recording that the block is stored is
Through the phase, available block is this time successfully found if condition is set up, program is returned, if do not found, cyclic variable k++, after
Continuous detection, jumps to step 3；
Step 6：NTmpBlock=0, i++ jumps to step 2.
3rd, when there is stream medium data to store into system, now no longer distribute storage file by VFS and deposited
Storage, but according to equipment distributing buffer computing formula, corresponding caching is distributed for headend equipment from caching system；So as to this sets
Standby data flow will always be written as its spatial cache specified.
4th, distribute memory block for the equipment of front end in hard drive space, after buffering area is write completely, then can be to depositing that it is specified
Storage block write data, now, should detect whether memory block be enough to accommodate current cache area data, if can, write direct；If no
Then next available memory block can be detected with available block probe algorithm, and distribute to the headend equipment, as currently stored piece.
5th, to the data in the region write buffering area of specified memory block in the way of asynchronous write, and this is buffered
Area adds busy buffer queue.Whether there is the buffering for completing asynchronous write in the busy buffer queue of detection, and returned by caching system
Receipts have completed the buffering area of asynchronous write, according to caching formula are again now that equipment distributes spatial cache, continue executing with storage.
This storage algorithm be applied to streaming media video storage tape come value have it is following some.
1. from the security of security protection industry for, because this algorithm has abandoned the storage mode of file system, with direct
Pattern opens disk, has got around disk cache, and so as to directly to the region that disk is specified be written and read operation, and each road regards
Frequency is allocated according to big minor matter in the buffer according to algorithm, so as to its size has stationarity and knowability, so as to
In the case of server exception, the loss of video data can be to greatest extent reduced, it is ensured that the security of monitoring data.
2. on the other hand, due to system by disk according to fixed size, several memory blocks are divided into（chunk）, often
Individual headend equipment can only write data to the memory block distributed for it, search according to available block probe algorithm after memory block is write completely
Available block, then stored from this block, so as to, effectively eliminate because hard disc data accesses the fragment problems for causing repeatedly, be
System performance is more stable, and hard disk service life increases, and hard disk utilization reaches 100%, in large-scale monitoring project, such as safety city
In the construction in city, this will carry out huge economic benefit for system background storage tape.
3. the present invention is using homemade spatial cache due to carrying out data buffer storage, thus its cycle for being write to hard disk
It is fixed, this causes hard disk groups are not required to whenever all be driven, and is simply periodically driven, this will makes storage
The power consumption of equipment is substantially reduced, and becomes energy-conservation green product.
4., because this storage algorithm avoids system server for the bottleneck for processing extensive video code flow, this algorithm can
So that the performance of storage server is greatly improved, the video code flow that separate unit storage server equipment can be processed can exceed
500Mbps。
5. because whole storing process is controlled by the algorithm of the present invention, according to table in many experiments and actual application environment
It is existing, its initial hit rate（Calculate for the first time in circulation）Reach more than 80%, i.e. the reading efficiency of hard disk increases, equipment performance is carried
Height, user interactivity is improved well.
Description of the drawings
Accompanying drawing 1 is the schematic diagram of the storing process of current operation system acquiescence.
Accompanying drawing 2 is that current operation system fragment in storage produces process schematic.
Accompanying drawing 3 is to use network video data storing process schematic diagram after inventive algorithm.
Accompanying drawing 4 is used after inventive algorithm without disk fragmentses reason schematic diagram.
Accompanying drawing 5 is the present invention whole workflow schematic diagram within the storage system.
Accompanying drawing 6 is the storage available block probe algorithm schematic flow sheet in present invention storage algorithm.
Specific embodiment
Technical scheme is further illustrated below in conjunction with the accompanying drawings.
Contrast accompanying drawing 1 and accompanying drawing 3, it can be seen that Internet video before storage is ready for into system, deposit by acquiescence
Storage mode is that video data is stored into into small documents one by one by Virtual File System, then again will by way of file
Video data writes into disk cache, finally irregularly data is write into into hard disk.The present invention is special for stream medium data
Property, whole data storage procedure is improved.
Referring to accompanying drawing 1, front network video data is ready for storage into system, is now directly entered virtual file system
System, and abandoned first by file come the way of data storage, subsequent data enter self-defining caching system（This caching by
Installed System Memory distributes）, and and non-entry system default disk cache, specific embodiment is described as follows.
1. caching system is distributed in most of space of Installed System Memory, according to Buddy system algorithm（Outside for solution system
Fragment and the algorithm that uses）Storage allocation is managed.
2. caching when being each headend equipment distribution write memory block with reference to the characteristics of stream medium data.Such as accompanying drawing 5, vacation
As front end there are two-way high-definition network video data to enter system, according to caching distribution formula, it is assumed that C takes 2, represent that video data can
To deposit in the buffer 2 seconds, if the desired value of the code stream of computing device 1 is 4Mbps, the desired value of the code stream of equipment 2 is 2Mbps,
Then system cache is 2*4=8Mb for the spatial cache that equipment 1 is provided, and the spatial cache for being the offer of equipment 2 is 2*2=4Mb, once
After initialization, as long as headend equipment video code flow is stable, then the spatial cache for distributing is held essentially constant, and now, equipment 1 and sets
Standby 2 data are respectively stored into each self-corresponding caching；After 2 seconds, the video data in each caching writes full, now drives hard
Data in caching are write into hard disk by disk group.
Get around system cache and be described as follows using the purpose of self-defined caching.
On the one hand, because the advantage of disk cache is for commonly used data, due to its be present in advance it is slow
Deposit, thus read or write speed is very fast, and in monitoring trade, video stream data is at storage state in most times, nothing
It need to be operated, now, due to memory space that disk cache is the distribution of each video equipment be it is uncontrollable,
I.e.：Certain section of video data may occur that to be in a large number stranded in caching internal, if now system cut-off or machine of delaying, this segment data is lost
Mistake may bring and have a strong impact on.And self-defining caching system, it is controllable that it is the memory space of each video equipment distribution
And it is knowable, be also not in a large amount of situations about losing of certain section of video even if power-off so.
On the other hand, because the memory space that disk cache is each video equipment distribution is uncontrollable, this also makes
Uncontrollable to the time of hard disk write data into caching, substantially hard disk groups need to drive always, and power consumption increases, hard disk life-span
Reduce；And self-defining caching system, it is fixed as the time that hard disk writes data, such as every write in 2 seconds once, thus
Hard disk is only needed to be driven every 2 seconds and once write.
3. as shown in Figure 5, step 1 experienced, after two equipment are write video data in caching system, when
When caching writes full, into step 2, hard disk is write data into, now hard disk is not according to the storage mode of acquiescence（Be segmented,
Fragmentation）, but by memory block that hard disk partition is several formed objects（chunk）, it is that equipment distributes two when initial
Memory block, such as：The video data of equipment 1 is stored in chunk0, and the video data of equipment 2 is stored in chunk5, following period of time thereafter,
The video data of two equipment is deposited into fixation in this memory block, until this memory block space exhausts.When the sky of this memory block
Between when being finished, into step 3（Accompanying drawing 5）, further according to storage detection block probe algorithm, the chunk that the next one is stored is found,
So constantly circulation；Detection mode is as follows（Referring to accompanying drawing 6）.
（1）Assume that hard disk is divided into 16 memory blocks, i.e. T=16, chunk0- chunk15, the video equipment of front end two
The mathematic expectaion of produced total code stream is 6Mbps, then proportion P1=upper of the code stream of two video equipments in total code stream
（6/4）=2, P1=upper（6/2）=3.
（2）If video equipment 1 has write full chunk0, calculated according to algorithm, possible available memory block nActBlock=
NTmpBlock % total block data T；nTmpBlock = a + (P – i ) * k；Then nTmpBlock=0+（2-0）* 1=2, tries to achieve
nActBlock=2%16=2；If now chunk2 is idle, or the video recording spatially is expired, then video equipment 2
Video stream data will be stored on chunk2；If chunk2 is occupied, the value of i and k is added 1, continues to calculate,
Till until finding idle interval.It is identical with 1 that the memory block of equipment 2 finds mode.
（3）During video data is deleted, as long as there is space to be freed out, just have an opportunity by other video datas
Storage is used, and is not in the situation that little space can not be utilized.
Video data does not adopt the segmentation paging system storage mode of system default, and adopts this kind of storage mode, its objective is.
Referring to the drawings 2, the segmentation paging system of system default is stored, due in storing process, what the memory space of hard disk was divided
The size of each memory space is differed, and in the finding algorithm according to the system, some smaller spaces will be likely to
Can not be utilized, or after the deletion repeatedly of video data, the little space that many can not be reused be will be left behind, so as to produce
Raw fragment（And the defragmentation mechanism for starting hard disk is time-consuming and cost source a method）, and pass through present invention storage algorithm
Storage video data（Referring to the drawings 4）, the storage location change of each video is little, only by current space storage it is full it
Afterwards, can just transfer to find next space and be stored, because stream medium data now is not stored with file mode,
Thus, either how little space, still can utilize, so as to so that hard drive space is completely absent fragment.Also, mention
Available block probe algorithm, according in many experiments and actual application environment show, its initial hit rate（Count for the first time in circulation
Calculate）More than 80% is reached, is reduced and is covered the fluctuation for writing situation writing speed repeatedly.
Without fragment store system efficient, stable operation for a long time, and hard disk utilization rate and service life are obtained
Greatly improve.Below by taking a safety monitoring project as an example, the value dimension of the invention is contrasted.
Headend equipment has 1000 road high-definition network cameras, and code stream is 4Mbps, it is desirable to which whole day is recorded a video, and number of recording a video
According to preservation 15 days.Following table is the performance comparison of two kinds of different storage modes.
| Performance | Default storage mode | Stream medium data stores algorithm |
| Storage server quantity（Platform） | 27 | 9 |
| Required hard disk quantity（3T） | 263 | 206 |
| Hard disk low-quality disk rate（/ year） | 30% | 5% |
| Power consumption（/ hour） | 9345W | 4890W |
The present invention can be used in any storage device in monitoring system, including DVR（Digital hard disc video recorder）、NVR
（Network hard disk video recorder）And other storage arrays etc.；The above is not limited to the present invention, all originals in the present invention
Then with spirit within any modification, equivalent and the improvement made etc., be included within protection scope of the present invention.
Claims (9)
1. it is a kind of to be adapted to the block storage algorithm applicable that multipath concurrence writes stream medium data, it is characterised in that for network
In video monitoring, optimization in accordance with the following steps stores the stream medium data of web camera：
Initialization caching system, it is corresponding slow for each headend equipment distribution then in conjunction with the stream medium data feature of headend equipment
Area is deposited, wherein, the headend equipment is web camera；
Initialization hard disk storage system, is then divided into several memory blocks by the disk of opening according to fixed size, and is
Each headend equipment specifies a memory block；
The stream medium data of each headend equipment is written in corresponding buffer area；
Certain headend equipment buffer area write it is full after, with asynchronous WriteMode by the stream medium data in the buffer area be written to
In the corresponding memory block of the headend equipment.
2. a kind of as claimed in claim 1 to be adapted to the block storage algorithm applicable that multipath concurrence writes stream medium data, it is special
Levy and be, the step of the stream medium data feature of the combination headend equipment distributes corresponding buffer area for each headend equipment
In, distribute buffer size Fb of headend equipment according to equation below：
Fb=C* E(Q)
In formula, C is the buffering number of seconds of caching system, and Q is code stream size of the stream medium data of the headend equipment within the unit second,
E (Q) is code stream mathematic expectaion of the stream medium data of the headend equipment within the unit second.
3. a kind of as claimed in claim 1 to be adapted to the block storage algorithm applicable that multipath concurrence writes stream medium data, it is special
Levy and be, it is before stream medium data in the buffer area is written in memory block corresponding with the headend equipment including as follows
Step：Whether corresponding with the headend equipment memory block of detection be enough to accommodate the stream medium data in the buffer area, if can not if
For the headend equipment an available memory block is reassigned as current memory block.
4. a kind of as claimed in claim 3 to be adapted to the block storage algorithm applicable that multipath concurrence writes stream medium data, it is special
Levy and be, it is described reassign an available memory block for the headend equipment the step of in comprise the steps：
Step 1. initializes following variable：Cyclic variable i=0, cyclic variable k=1, temporary variable nTmpBlock=0；
Whether step 2. judges the value of cyclic variable i less than proportion P of the headend equipment code stream in total code stream, if condition is set up
Continue downward execution step 3 and 4, the proportion P=upper (E (S)/E (Q)), in formula, Q is the Streaming Media of the headend equipment
Code stream size of the data within the unit second, E (Q) is code stream mathematics phase of the stream medium data of the headend equipment within the unit second
Hope, S is total code stream size of the stream medium data of all headend equipments within the unit second, and E (S) is the stream matchmaker of all headend equipments
Total code stream mathematic expectaion of the volume data within the unit second；
Step 3. asks the numbering nActBlock=nTmpBlock % T of next possible available memory block, in formula, temporary variable
NTmpBlock=a+(P-i) * k, a are the currently stored block number specified, and T is memory block in hard disk storage system
Sum；
Whether step 4. judges temporary variable nTmpBlock less than currently stored block number a and block storage sum T sums, if bar
Part is false and jumps to step 6, if condition is set up continues downward execution step 5；
Step 5. judges whether memory block corresponding with numbering nActBlock is not used or the storage by remaining headend equipment
The stream medium data that block is stored is expired, and available memory block is successfully found if condition is set up, and program is returned, if do not had
Find, cyclic variable k++, continue to detect, jump to step 3；
Step 6. temporary variable nTmpBlock=0, i++ jumps to step 2.
5. a kind of as claimed in claim 1 to be adapted to the block storage algorithm applicable that multipath concurrence writes stream medium data, it is special
Levy and be, stream medium data in the buffer area is written to by storage corresponding with the headend equipment with asynchronous WriteMode described
While in block, the buffer area is added in busy buffer area queue, then in conjunction with the stream medium data feature of the headend equipment
Again corresponding buffer area is distributed for the headend equipment.
6. a kind of as claimed in claim 5 to be adapted to the block storage algorithm applicable that multipath concurrence writes stream medium data, it is special
Levy and be, detect with the presence or absence of the buffer area for completing asynchronous write in the busy buffer area queue, if existing, returned by caching system
Receive the buffer area.
7. a kind of as claimed in claim 1 to be adapted to the block storage algorithm applicable that multipath concurrence writes stream medium data, it is special
Levy and be, the stream medium data of each headend equipment is written in corresponding buffer area, by week with multipath concurrence writing mode
Phase and the stream medium data in buffer area is written in corresponding memory block with single thread multi-path asynchronous writing mode.
8. a kind of as claimed in claim 7 to be adapted to the block storage algorithm applicable that multipath concurrence writes stream medium data, it is special
Levy and be, the cycle is controllable.
9. a kind of as claimed in claim 1 to be adapted to the block storage algorithm applicable that multipath concurrence writes stream medium data, it is special
Levy and be, the buffer area for distributing is managed according to Buddy system algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310394300.8A CN103702057B (en) | 2013-09-03 | 2013-09-03 | Block storage algorithm applicable to multiple paths of concurrent-written stream media data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310394300.8A CN103702057B (en) | 2013-09-03 | 2013-09-03 | Block storage algorithm applicable to multiple paths of concurrent-written stream media data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103702057A CN103702057A (en) | 2014-04-02 |
| CN103702057B true CN103702057B (en) | 2017-04-12 |
Family
ID=50363444
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310394300.8A Active CN103702057B (en) | 2013-09-03 | 2013-09-03 | Block storage algorithm applicable to multiple paths of concurrent-written stream media data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103702057B (en) |
Families Citing this family (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104159058A (en) * | 2014-07-26 | 2014-11-19 | 佳都新太科技股份有限公司 | Efficient method capable of realizing simultaneous video recording of multi-way media streams |
| CN104156173A (en) * | 2014-07-29 | 2014-11-19 | 浪潮(北京)电子信息产业有限公司 | Method and device for disk defragmentation |
| CN104811647B (en) * | 2015-05-15 | 2018-11-06 | 电子科技大学 | The double subregion wiring methods of distributed memory system disk towards video stream data |
| CN104811646B (en) * | 2015-05-15 | 2018-07-24 | 电子科技大学 | The storage method of the modulation of multiple video strems Data Concurrent and buffering based on Coutinuous store model |
| CN107615253B (en) | 2015-06-01 | 2021-03-12 | 深圳市大疆创新科技有限公司 | System and method for memory architecture |
| CN105306865B (en) * | 2015-10-29 | 2019-05-17 | 四川奇迹云科技有限公司 | A kind of video monitoring data system |
| CN105389126B (en) * | 2015-10-29 | 2019-02-15 | 四川奇迹云科技有限公司 | A kind of block storage system of video monitoring data |
| CN105357463B (en) * | 2015-10-29 | 2019-01-08 | 北京易视云科技有限公司 | A kind of storage method of video monitoring data |
| CN105451028B (en) * | 2015-12-03 | 2019-04-02 | 深圳市泛海三江电子股份有限公司 | Storage server and implementation method thereof |
| CN106293518B (en) * | 2016-07-29 | 2019-09-10 | 成都神州数码索贝科技有限公司 | A kind of efficient security industry media data storage file system |
| CN107333080B (en) * | 2017-05-24 | 2019-11-22 | 清华大学 | A kind of high speed, high resolution image data recording method, apparatus and system |
| CN110737389B (en) * | 2018-07-19 | 2023-05-16 | 杭州海康威视系统技术有限公司 | Method and device for storing data |
| CN109582521B (en) * | 2018-12-10 | 2022-04-29 | 浪潮(北京)电子信息产业有限公司 | Method, device, equipment and medium for testing read-write performance of storage system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101742263A (en) * | 2009-12-08 | 2010-06-16 | 北京互信互通信息技术股份有限公司 | Method for storing surveillance video data |
| CN103186349A (en) * | 2011-12-27 | 2013-07-03 | 杭州信核数据科技有限公司 | Block-level distributed storage system and data reading-writing method thereof |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8219750B2 (en) * | 2008-06-30 | 2012-07-10 | Pivot3 | Method and system for execution of applications in conjunction with distributed RAID |
-
2013
- 2013-09-03 CN CN201310394300.8A patent/CN103702057B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101742263A (en) * | 2009-12-08 | 2010-06-16 | 北京互信互通信息技术股份有限公司 | Method for storing surveillance video data |
| CN103186349A (en) * | 2011-12-27 | 2013-07-03 | 杭州信核数据科技有限公司 | Block-level distributed storage system and data reading-writing method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103702057A (en) | 2014-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103702057B (en) | Block storage algorithm applicable to multiple paths of concurrent-written stream media data | |
| US9665291B2 (en) | Migration of hot and cold data between high-performance storage and low-performance storage at block and sub-block granularities | |
| CN102843396B (en) | Data write-in and read method and device in a kind of distributed cache system | |
| US8654472B2 (en) | Implementing enhanced fragmented stream handling in a shingled disk drive | |
| US8590001B2 (en) | Network storage system with data prefetch and method of operation thereof | |
| CN101981551B (en) | Apparatus and method for cache utilization | |
| US8291245B2 (en) | Method, apparatus and system for reducing power consumption based on storage device data migration | |
| US20060112232A1 (en) | Method and apparatus for managing a cache memory in a mass-storage system | |
| US20210096777A1 (en) | Method for predicting lba information, and ssd | |
| CN101644996A (en) | Storage method of index data and storage control device | |
| EP3425513B1 (en) | Cache management method and apparatus for non-volatile storage device | |
| CN103279429A (en) | Application-aware distributed global shared cache partition method | |
| CN104503703B (en) | The treating method and apparatus of caching | |
| CN102999428A (en) | Four-stage addressing method for tile recording disk | |
| US20110225369A1 (en) | Multiport data cache apparatus and method of controlling the same | |
| CN107168651A (en) | A kind of small documents polymerize storage processing method | |
| CN104811646B (en) | The storage method of the modulation of multiple video strems Data Concurrent and buffering based on Coutinuous store model | |
| KR20100069240A (en) | Apparatus and method for cache control | |
| CN107506146A (en) | A kind of data-storage system | |
| US7984313B2 (en) | Method, apparatus and system for reducing power consumption involving data storage devices | |
| CN109375868B (en) | Data storage method, scheduling device, system, equipment and storage medium | |
| CN101957729B (en) | Logical block transformation method and method and device compatible with reading and writing of user based on same | |
| CN105389126A (en) | Blocking storage system of video monitoring data | |
| CN112685337B (en) | Method for hierarchically caching read and write data in storage cluster | |
| CN102779098A (en) | Hybrid caching cooperative adaptive prefetching method, middleware and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20170531 Address after: 610000 2, building 10, No. 2, 10, two, 2 Garden Road, Chengdu hi tech Zone, Sichuan, China Patentee after: CHENGDU XINYU TECHNOLOGY CO., LTD. Address before: The middle high tech Zone Yizhou road in Chengdu city of Sichuan province 610000 1800 No. 1 building 7 room 702 Patentee before: JUN TAI TECHNOLOGY CO., LTD. OF CHENGDU |