CN109343789A - A kind of reading accelerated method, device and electronic equipment based on IO scene Recognition - Google Patents

A kind of reading accelerated method, device and electronic equipment based on IO scene Recognition Download PDF

Info

Publication number
CN109343789A
CN109343789A CN201810866060.XA CN201810866060A CN109343789A CN 109343789 A CN109343789 A CN 109343789A CN 201810866060 A CN201810866060 A CN 201810866060A CN 109343789 A CN109343789 A CN 109343789A
Authority
CN
China
Prior art keywords
scene
read requests
grouping
read
library
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.)
Granted
Application number
CN201810866060.XA
Other languages
Chinese (zh)
Other versions
CN109343789B (en
Inventor
陈元强
李文祥
吴建辉
吕定灿
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen mulangyun Technology Co.,Ltd.
Original Assignee
Shenzhen Mulangyun Data Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenzhen Mulangyun Data Co Ltd filed Critical Shenzhen Mulangyun Data Co Ltd
Publication of CN109343789A publication Critical patent/CN109343789A/en
Application granted granted Critical
Publication of CN109343789B publication Critical patent/CN109343789B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
    • G06F3/061—Improving I/O performance
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0655—Vertical data movement, i.e. input-output transfer; data movement between one or more hosts and one or more storage devices
    • G06F3/0656—Data buffering arrangements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
    • G06F3/067—Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Memory System Of A Hierarchy Structure (AREA)
  • Signal Processing For Digital Recording And Reproducing (AREA)

Abstract

The embodiment of the present application discloses a kind of reading accelerated method based on IO scene Recognition, is applied to distributed memory system, which comprises receives I O read requests；The scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining IO scene grouping feature library corresponding with the IO scene of the scene matching of I O read requests grouping library；Starting IO scene is pre-read, and will be grouped the corresponding IO scene grouping feature in library with the IO scene of the scene matching of the I O read requests and the corresponding data in magnetic disk of the I O read requests reads memory cache in advance；The I O read requests are responded, the memory cache pre-read is read out, return to the corresponding data of the I O read requests.By this method, IO may be implemented and read to accelerate.

Description

A kind of reading accelerated method, device and electronic equipment based on IO scene Recognition
Technical field
This application involves field of cloud computer technology, more particularly, to a kind of reading accelerated method based on IO scene Recognition, Device and electronic equipment.
Background technique
Under cloud computing environment, when reading and writing data using distributed memory system, usually as unit of data block, data block From KB to MB etc..One file or chunk data are written in distributed memory system, usually to be divided again data block Block, and random distribution is into the disk of each server.Since standard machinery disk read-write speed is usually slow, especially with Machine-readable, if data block is in KB rank, each second is also with regard to several MB.
Summary of the invention
In view of the above problems, present applicant proposes a kind of reading accelerated method, device and electronics based on IO scene Recognition to set It is standby, solve above-mentioned technical problem.
The embodiment of the present application provides a kind of reading accelerated method based on IO scene Recognition, is applied to distributed storage system System, which comprises receive I O read requests；The scene of the I O read requests and IO scene grouping set are subjected to similarity Match, determining IO scene grouping feature library corresponding with the IO scene of the scene matching of I O read requests grouping library；Starting IO Scape is pre-read, and will be grouped the corresponding IO scene grouping feature library in library and the IO with the IO scene of the scene matching of the I O read requests The corresponding data in magnetic disk of read request reads memory cache in advance；The I O read requests are responded, the memory cache pre-read is read out, Return to the corresponding data of the I O read requests.
The embodiment of the present application also provides a kind of reading accelerator based on IO scene Recognition, described device includes: to receive Module, processing module, pre- read through model and respond module, wherein the receiving module is for receiving I O read requests；The processing Module is used to the scene of the I O read requests and IO scene grouping set carrying out similarity mode, the determining and I O read requests Scene matching the corresponding IO scene grouping feature library in IO scene grouping library；The pre- read through model is pre- for starting IO scene It reads, the corresponding IO scene grouping feature library in library will be grouped with the IO scene of the scene matching of the I O read requests and IO reading is asked Corresponding data in magnetic disk is asked to read memory cache in advance；The corresponding module is for responding the I O read requests, to the memory pre-read Caching is read out, and returns to the corresponding data of the I O read requests.
The embodiment of the present application also provides a kind of, and the reading based on IO scene Recognition accelerates electronic equipment, the electronic equipment packet Include memory and processor, the memory is couple to the processor, the memory store instruction, when described instruction by The processor executes the reading accelerated method based on IO scene Recognition that above-mentioned first aspect provides when processor execution.
Compared with the existing technology, reading accelerated method, device and the electronic equipment provided by the present application based on IO scene Recognition, By receiving I O read requests, the scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining and institute The IO scene grouping library of the scene matching of I O read requests is stated, starting IO scene is pre-read, by the scene matching with the I O read requests The IO scene grouping corresponding IO scene grouping feature library in library and the corresponding data in magnetic disk of the I O read requests to read memory in advance slow It deposits, responds the I O read requests, the memory cache pre-read is read out, return to data corresponding with the I O read requests, from And it realizes IO and reads to accelerate.
These aspects or other aspects of the application can more straightforward in the following description.
Detailed description of the invention
In order to more clearly explain the technical solutions in the embodiments of the present application, make required in being described below to embodiment Attached drawing is briefly described, it should be apparent that, the drawings in the following description are only some examples of the present application, for For those skilled in the art, without creative efforts, it can also be obtained according to these attached drawings other attached Figure.
Fig. 1 shows the reading accelerated method flow chart based on IO scene Recognition of the application first embodiment proposition.
Fig. 2 shows the reading accelerated method flow charts based on IO scene Recognition that the application second embodiment proposes.
Fig. 3 shows the reading accelerated method flow chart based on IO scene Recognition of the application 3rd embodiment proposition.
Fig. 4 shows the reading accelerated method flow chart based on IO scene Recognition of the application fourth embodiment proposition.
Fig. 5 shows the reading accelerated method flow chart based on IO scene Recognition of the 5th embodiment of the application proposition.
Fig. 6 shows the reading accelerated method flow chart based on IO scene Recognition of the application sixth embodiment proposition.
Fig. 7 shows the reading accelerated method flow chart based on IO scene Recognition of the 7th embodiment of the application proposition.
Fig. 8 shows the reading accelerated method flow chart based on IO scene Recognition of the 8th embodiment of the application proposition.
Fig. 9 shows the reading accelerated method flow chart based on IO scene Recognition of the 9th embodiment of the application proposition.
Figure 10 shows the reading accelerated method flow chart based on IO scene Recognition of the tenth embodiment of the application proposition.
Figure 11 shows the reading accelerated method flow chart based on IO scene Recognition of the 11st embodiment of the application proposition.
Figure 12 shows the reading accelerated method flow chart based on IO scene Recognition of the 12nd embodiment of the application proposition.
Figure 13 shows a kind of knot of the reading accelerator based on IO scene Recognition of the 13rd embodiment of the application offer Structure block diagram.
Figure 14 show the embodiment of the present application for executing the reading based on IO scene Recognition according to the embodiment of the present application The block diagram of the electronic equipment of accelerated method.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present application, technical solutions in the embodiments of the present application carries out clear, complete Site preparation description, it is clear that described embodiments are only a part of embodiments of the present application, instead of all the embodiments.It is based on Embodiment in the application, it is obtained by those of ordinary skill in the art without making creative efforts every other Embodiment shall fall in the protection scope of this application.
Currently, when reading and writing data using distributed memory system, when application layer reads and writes a file, it will usually initiate tens of Ten thousand random I O read requests read and write a upper GB grades of data to disk, and speed can be comparable slow.Inventors have found that in distribution In system, (1) either random IO read operation or sequence IO read operation, the speed of reading are all slow；(2) it simply pre-reads, More data can be read, more networks, disk read-write performance consumption are brought；(3) for a large amount of regular discontinuous every time Random read-write, such as activation system and application, traversal file system uniform velocity can be very slow.Therefore, this hair is inventors herein proposed The reading accelerated method based on IO scene Recognition in bright embodiment.Each implementation in the application is specifically described below in conjunction with attached drawing Example.
First embodiment
Referring to Fig. 1, Fig. 1 shows the reading accelerated method based on IO scene Recognition of the application first embodiment offer Flow diagram.It will be explained in detail below for process shown in FIG. 1, the above-mentioned reading based on IO scene Recognition accelerates Method specifically may comprise steps of:
Step S110: I O read requests are received.
When storage bottom receives upper layer I O read requests, show that upper layer IO occurs and reads demand, the upper layer can be centre Layer, and/or application layer.
Step S120: the scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining and institute State the corresponding IO scene grouping feature library in IO scene grouping library of the scene matching of I O read requests.
The IO scene grouping set is the scene grouping set in IO historical data base, and has Continuity Analysis spy Sign.The scene of the I O read requests and IO scene grouping set are subjected to similarity mode, i.e., by the scene of the I O read requests Similarity mode is carried out with the scene grouping set in IO historical data base.
Similarity mode is the scene and the phase of the scene grouping set in IO historical data base of the I O read requests Like degree, find out in IO historical data base with the grouping in the immediate scene grouping set of the scene similarity of the I O read requests Library.Feature of the similarity closer to the IO scene grouping set shown in the scene characteristic and historical data base of the I O read requests Closer, i.e., the described IO scene grouping set and the scene similarity of the I O read requests are matched.
Determining IO scene grouping feature library corresponding with the IO scene of the scene matching of I O read requests grouping library, i.e., really It is fixed it is immediate with the scene characteristic of the I O read requests include that be grouped library corresponding for IO scene in IO scene grouping set IO scene grouping feature library filters out IO corresponding with the scene similarity of the I O read requests unmatched IO scene grouping library Jing Fenzutezhengku.
Step S130: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
IO scene point corresponding with the IO scene of the scene matching of I O read requests grouping library is determined in the step s 120 After group feature database, starting pre-reads scene, is included in IO history number for what the scene similarity with the I O read requests matched The corresponding IO scene grouping feature library in library is grouped according to the IO scene in library and the corresponding data in magnetic disk of the I O read requests is read in advance In memory cache.The corresponding data in magnetic disk of the I O read requests is the data in magnetic disk for including the I O read requests.
Step S140: responding the I O read requests, is read out to the memory cache pre-read, returns to the I O read requests pair The data answered.
Storage bottom responds the I O read requests, is read out to the memory cache pre-read, i.e., in memory cache with institute State the scene similarity mode of I O read requests includes the corresponding IO grouping in IO scene grouping library in IO scene grouping set Scene characteristic library and the corresponding data in magnetic disk of I O read requests are read out, and return to read IO scene grouping feature library and packet Data in magnetic disk containing I O read requests.
Second embodiment
Referring to Fig. 2, Fig. 2 shows the reading accelerated methods based on IO scene Recognition of the application second embodiment offer Flow diagram.It will be explained in detail below for process shown in Fig. 2, the reading acceleration side based on IO scene Recognition Method can specifically include following steps:
Step S210: I O read requests are received.
The detailed process of step S210 is referred to the step S110 in first embodiment, and which is not described herein again.
Step S220: initialization time window t1。
When storage bottom receives upper layer I O read requests, initialization time window t1, the initialization time window is one Empirical value.
Step S230: the time window t is counted1The scene characteristic of the interior I O read requests.
Count the time window t1The scene characteristic of the interior I O read requests, the scene characteristic are to read offset as base Plinth carries out the feature for discontinuously judging to obtain to set of records ends, and the scene characteristic includes file destination ID, and the time reads offset, Data block fingerprint and/or data block location.
Step S240: the IO scene grouping library in the scene characteristic of the I O read requests and IO scene grouping set is carried out Matching.
It is by the matching that the scene characteristic of the I O read requests is carried out with the IO scene grouping library in IO scene grouping set Similarity mode.
Step S250: it is special to obtain IO scene grouping corresponding with the IO scene of the scene matching of I O read requests grouping library Levy library.
After the similarity mode of step S240, obtain and the IO scene of the scene similarity mode of the I O read requests point The corresponding IO scene grouping feature library in group library.
Step S260: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
Step S270: responding the I O read requests, is read out to the memory cache pre-read, returns to the I O read requests pair The data answered.
The detailed process of step S260 to step S270 are referred to the step S130 to S140 in first embodiment, here It repeats no more.
The reading accelerated method based on IO scene Recognition that the application second embodiment provides, passes through statistical time window t1It is interior The scene characteristic of the I O read requests, and the IO scene in the scene characteristic of the I O read requests and IO scene grouping set is divided Group library with matched, obtain and be grouped library with the IO scene of the scene matchings of the I O read requests, start IO scene and pre-read, will be with The IO scene grouping corresponding IO scene grouping feature library in library of the scene matching of the I O read requests and the I O read requests are corresponding Data in magnetic disk read memory cache in advance, respond the I O read requests, the memory cache pre-read be read out, return with it is described The corresponding data of I O read requests.
3rd embodiment
Referring to Fig. 3, Fig. 3 shows the reading accelerated method based on IO scene Recognition of the application 3rd embodiment offer Flow diagram.It will be explained in detail below for process shown in Fig. 3, the reading acceleration side based on IO scene Recognition Method can specifically include following steps:
Step S310: I O read requests are received.
The detailed process of step S310 is referred to the step S210 in second embodiment, and which is not described herein again.
Step S320: time window t1It is initialized as the N% of analysis window t.
The analysis window t be IO historical data base establishment process in statistical history I O read requests scene characteristic when Between, by the time window t1It is initialized as the N% of analysis window t, i.e. t1=t × N%, N < 100, N value is bigger, the time Window t1It is bigger, then in the time window t1Interior I O read requests number is more, and the scene of I O read requests is more, then corresponds to I O read requests scene characteristic it is abundanter, then the scene characteristic of the I O read requests counted in subsequent step S330 It is more accurate.Correspondingly, N value is bigger, in biggish time window t1Interior more I O read requests number just will affect reading and accelerate Effect.
Step S330: the time window t is counted1The scene characteristic of the interior I O read requests.
Step S340: the IO scene grouping library in the scene characteristic of the I O read requests and IO scene grouping set is carried out Matching.
Step S350: it is special to obtain IO scene grouping corresponding with the IO scene of the scene matching of I O read requests grouping library Levy library.
Step S360: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
Step S370: responding the I O read requests, is read out to the memory cache pre-read, returns to the I O read requests pair The data answered.
The detailed process of step S340 to step S370 are referred to the step S240 to S270 in second embodiment, here It repeats no more.
The reading accelerated method based on IO scene Recognition that the application 3rd embodiment provides receives I/O Request, time window t1It is initialized as the N% of analysis window t, passes through statistical time window t1The scene characteristic of the interior I O read requests, and by the IO The scene characteristic of read request in IO scene grouping set IO scene grouping library with matched, obtain and the I O read requests The IO scene of scene matching be grouped library, starting IO scene is pre-read, by the IO scene of the scene matching with the I O read requests point The group corresponding IO scene grouping feature library in library and the corresponding data in magnetic disk of the I O read requests read memory cache in advance, described in response I O read requests are read out the memory cache pre-read, return to data corresponding with the I O read requests.
Fourth embodiment
Referring to Fig. 4, Fig. 4 shows the reading accelerated method based on IO scene Recognition of the application fourth embodiment offer Flow diagram.It will be explained in detail below for process shown in Fig. 4, the reading acceleration side based on IO scene Recognition Method can specifically include following steps:
Step S410: I O read requests are received.
Step S420: initialization time window t1。
Step S430: the time window t is counted1The scene characteristic of the interior I O read requests.
The detailed process of step S410 to step S430 are referred to the step S210 in second embodiment to step S230, Which is not described herein again.
Step S440: scanning the IO scene grouping set, by the scene characteristic of the I O read requests and the IO scene IO scene grouping library in grouping set carries out similarity mode.
The IO scene grouping set, i.e. IO scene grouping set in scanning IO historical data base are scanned, by step The time window t counted in S4301In the scene characteristic of the interior I O read requests and the IO scene grouping set IO scene is grouped library and carries out similarity mode, that is, calculates the similarity of the two, compares the similarity of the two.
Step S450: determining IO corresponding with the IO scene of the scene characteristic successful match of I O read requests grouping library Jing Fenzutezhengku.
The determining IO scene with the scene characteristic successful match of the I O read requests is grouped library, that is, passes through institute in step S440 The similarity of calculating, it is determining it is immediate with the scene characteristics of the I O read requests include the field IO in IO scene grouping set Scape is grouped the corresponding IO scene grouping feature library in library.
Step S460: it is special to obtain IO scene grouping corresponding with the IO scene of the scene matching of I O read requests grouping library Levy library.
Step S470: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
Step S480: responding the I O read requests, is read out to the memory cache pre-read, returns and the I O read requests Corresponding data.
The detailed process of step S460 to step S470 are referred to the step S240 to S270 in second embodiment, here It repeats no more.
The reading accelerated method based on IO scene Recognition that the application fourth embodiment provides receives I/O Request, when initialization Between window t1, pass through statistical time window t1The scene characteristic of the interior I O read requests scans the IO scene grouping set, will IO scene grouping library in the scene characteristic of the I O read requests and the IO scene grouping set carries out similarity mode, determines It is grouped library with the IO scene of the scene characteristic successful match of the I O read requests, is obtained and the scene matching of the I O read requests IO scene is grouped library, and starting IO scene is pre-read, by IO corresponding with the IO scene of the scene matching of I O read requests grouping library Scene grouping feature library and the corresponding data in magnetic disk of the I O read requests read memory cache in advance, respond the I O read requests, right The memory cache pre-read is read out, and returns to data corresponding with the I O read requests.
5th embodiment
Referring to Fig. 5, Fig. 5 shows the reading accelerated method based on IO scene Recognition of the 5th embodiment of the application offer Flow diagram.It will be explained in detail below for process shown in fig. 5, the reading acceleration side based on IO scene Recognition Method can specifically include following steps:
Step S501: I O read requests are received.
Step S502: time window t1It is initialized as the N% of analysis window t.
Step S503: the time window t is counted1The scene characteristic of the interior I O read requests.
The detailed process of step S501 to step S503 are referred to the step S410 in fourth embodiment to step S430, Which is not described herein again.
Step S504: setting number tc, count the average offset value a of the IO scene grouping set0。
The number tcFor empirical value, the number t is setc, count the average offset value a of the IO scene grouping set0 Calculate the average offset value of the IO scene grouping set.
Step S505: based on first IO offset, in the continuous number tcIt is interior, it obtains the I O read requests and exists The time window t1Interior offset average value a1, calculate a1With the average offset value a0The difference subtracted each other, then divided by described average Deviant a0Quotient, the quotient and 0 size.
It is described deviated by first IO based on, in the continuous number tcIt is interior, the I O read requests are obtained described Time window t1Interior offset average value a1, calculate a1With the average offset value a0The difference subtracted each other, then divided by the mean deviation Value a0Quotient calculating process, that is, similarity mode process.
Step S506: if the quotient is equal to 0, determine that the scene of the I O read requests and the IO scene are grouped library With success, then using the corresponding deviant of first IO of the I/O Request as indexing parameter.
If the quotient is equal to 0, show the scene characteristic of the I O read requests and the feature in IO scene grouping library Similarity is close, that is, determines the scene and IO scene grouping storehouse matching success of the I O read requests, then by the I/O Request The corresponding deviant of first IO as indexing parameter, since the indexing parameter is a deviant, by described Indexing parameter, which can be positioned quickly, is grouped planting modes on sink characteristic with the IO scene of the field characteristic matching of the I/O Request.
Step S507: if the quotient is not equal to 0, continue in next continuous number tcInterior calculating, until It can not find matching offset.
If the quotient is not equal to 0, show the scene characteristic of the I O read requests and the spy in IO scene grouping library Sign similarity is very low, that is, determines that the scene of the I O read requests and IO scene grouping storehouse matching are unsuccessful.Then continue next A continuous number tcInterior similarity calculation, until by the I O read requests traverse to can not find matching offset.
Step S508: determining IO corresponding with the IO scene of the scene characteristic successful match of I O read requests grouping library Jing Fenzutezhengku.
Step S509: it is special to obtain IO scene grouping corresponding with the IO scene of the scene matching of I O read requests grouping library Levy library.
Step S510: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
Step S511: responding the I O read requests, is read out to the memory cache pre-read, returns and the I O read requests Corresponding data.
The detailed process of step S508 to step S511 are referred to the step S450 to S480 in fourth embodiment, here It repeats no more.
The reading accelerated method based on IO scene Recognition that the 5th embodiment of the application provides receives I/O Request, time window t1It is initialized as the N% of analysis window t, statistical time window t1The scene characteristic of the interior I O read requests sets number tc, system Count the average offset value a of the IO scene grouping set0, based on first IO offset, in the continuous number tcIt is interior, The I O read requests are obtained in the time window t1Interior offset average value a1, calculate a1With the average offset value a0Subtract each other Difference, then divided by the average offset value a0Quotient, the quotient and 0 size, if the quotient be equal to 0, sentence The scene and IO scene grouping storehouse matching success of the fixed I O read requests, then it is first IO of the I/O Request is corresponding Deviant is as indexing parameter, if the quotient is not equal to 0, continues in next continuous number tcInterior calculating, directly Offset is matched to can not find, the determining IO scene with the scene characteristic successful match of the I O read requests is grouped library, acquisition and institute The IO scene grouping library of the scene matching of I O read requests is stated, starting IO scene is pre-read, by the scene matching with the I O read requests The IO scene grouping corresponding IO scene grouping feature library in library and the corresponding data in magnetic disk of the I O read requests to read memory in advance slow It deposits, responds the I O read requests, the memory cache pre-read is read out, return to data corresponding with the I O read requests.
Sixth embodiment
Referring to Fig. 6, Fig. 6 shows the reading accelerated method based on IO scene Recognition of the application sixth embodiment offer Flow diagram.It will be explained in detail below for process shown in fig. 6, the reading acceleration side based on IO scene Recognition Method can specifically include following steps:
Step S610: I O read requests are received.
Step S620: the scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining and institute State the corresponding IO scene grouping feature library in IO scene grouping library of the scene matching of I O read requests.
The detailed process of step S610 to step S620 are referred to the step S110 to S120 in first embodiment, here It repeats no more.
Step S630: starting IO scene is pre-read.
The scene of the I O read requests and IO scene grouping set are subjected to similarity mode, the determining and I O read requests Scene matching IO scene grouping library after, starting IO scene pre-read.
Step S640: by IO scene grouping feature corresponding with the IO scene of the scene matching of I O read requests grouping library Library and the corresponding data in magnetic disk of the I O read requests are saved in memory cache.
IO scene grouping feature library corresponding with the IO scene of the scene matching of I O read requests grouping library is positioned, and will The corresponding IO scene grouping feature library in library and the I O read requests pair are grouped with the IO scene of the scene matching of the I O read requests The data in magnetic disk answered reads memory cache in advance.
Step S650: setting time K, when in the setting time K to the memory cache without access, discharge in described Deposit caching.
Setting time K, the time K is empirical value, such as: K≤5 minute can be set.When in the setting time K To the memory cache without access, show the demand currently without I O read requests, discharges the memory cache.
Step S660: responding the I O read requests, is read out to the memory cache pre-read, returns and the I O read requests Corresponding data.
The detailed process of step S660 is referred to the step S140 in first embodiment, and which is not described herein again.
The reading accelerated method based on IO scene Recognition that the application sixth embodiment provides receives I/O Request, by the IO The scene and IO scene grouping set of read request carry out similarity mode, the determining field IO with the scene matching of the I O read requests Scape is grouped library, and starting IO scene is pre-read, and starting IO scene is pre-read, and the IO scene of the scene matching with the I O read requests is grouped The corresponding IO scene grouping feature library in library and the I O read requests and the corresponding data in magnetic disk of the I O read requests are saved in memory Caching, setting time K, when in the setting time K to the memory cache without access, discharge the memory cache, respond The I O read requests are read out the memory cache pre-read, return to data corresponding with the I O read requests.
7th embodiment
Referring to Fig. 7, Fig. 7 shows the reading accelerated method based on IO scene Recognition of the 7th embodiment of the application offer Flow diagram.It will be explained in detail below for process shown in Fig. 7, the reading acceleration side based on IO scene Recognition Method can specifically include following steps:
Step S710: I O read requests are received.
Step S720: the scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining and institute State the corresponding IO scene grouping feature library in IO scene grouping library of the scene matching of I O read requests.
Step S730: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
The detailed process of step S710 to step S730 are referred to the step S110 to S130 in first embodiment, here It repeats no more.
Step S740: access includes IO scene point corresponding with the IO scene of the scene matching of I O read requests grouping library The memory cache of group feature database and the corresponding data in magnetic disk of the I O read requests.
It is pre-read in starting IO scene, by IO scene corresponding with the IO scene of the scene matching of I O read requests grouping library After grouping feature library and the corresponding data in magnetic disk of the I O read requests read memory cache in advance, access with IO reading comprising asking The memory cache in the IO scene grouping library for the scene matching asked.
Step S750: if there is no the fields IO with the scene matching of the I O read requests in the memory cache of the access Scape is grouped library, then continues to initiate read requests to memory node.
If being not present in the memory cache of the access and being grouped library with the IO scene of the scene matching of the I O read requests, Illustrate the IO scene characteristic in the IO scene grouping library in IO scene grouping set and the scene characteristic similarity of the I O read requests Low namely similarity mode is unsuccessful, is not the content to be read of the I O read requests, therefore does not need to read in memory cache Data then continue to initiate read requests to memory node.
Step S760: if there is the IO scene with the scene matching of the I O read requests in the memory cache of the access It is grouped library, then returns to the corresponding data of the I O read requests.
If existing in the memory cache of the access and being grouped library with the IO scene of the scene matching of the I O read requests, say The IO scene characteristic and the scene characteristic similarity of the I O read requests in the IO scene grouping library in bright IO scene grouping set connect Closely namely characteristic similarity successful match, it is therefore desirable to read the data in memory cache, and it is corresponding to return to the I O read requests Data.
The reading accelerated method based on IO scene Recognition that the 7th embodiment of the application provides receives I O read requests, will be described The scene and IO scene grouping set of I O read requests carry out similarity mode, the determining IO with the scene matching of the I O read requests Scene is grouped the corresponding IO scene grouping feature library in library, and starting IO scene is pre-read, by the scene matching with the I O read requests The IO scene grouping corresponding IO scene grouping feature library in library and the corresponding data in magnetic disk of the I O read requests read memory cache in advance, Access is read comprising the scene grouping feature library IO and the IO answered with the IO scene of the scene matching of I O read requests grouping library The memory cache of corresponding data in magnetic disk is requested, if there is no the fields with the I O read requests in the memory cache of the access The matched IO scene of scape is grouped library, then continues to initiate read requests to memory node, if deposited in the memory cache of the access It is grouped library in the IO scene of the scene matching with the I O read requests, then returns to the corresponding data of the I O read requests.
8th embodiment
Referring to Fig. 8, Fig. 8 shows the reading accelerated method based on IO scene Recognition of the 8th embodiment of the application offer Flow diagram.It will be explained in detail below for process shown in Fig. 8, the reading acceleration side based on IO scene Recognition Method can specifically include following steps:
Step S810: the record associated IO information of file destination.
Before receiving I O read requests, the record associated IO information of file destination needs that is, before receiving I O read requests Establish history I/O data library.The associated IO information of record file destination includes file destination ID, the time, reads offset, data Block fingerprint and data block location.
Step S820: the IO that analysis obtains the file destination reads scene behavior, obtains IO scene grouping set.
After recording the associated IO information of file destination, the IO that analysis obtains the file destination reads scene behavior, i.e., Discontinuous discriminatory analysis is carried out to the set of records ends of the file destination based on reading offset and obtains the IO of the file destination Read scene behavior.
Step S830: I O read requests are received.
Step S840: the scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining and institute State the IO scene grouping library of the scene matching of I O read requests；
Step S850: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance；
Step S860: responding the I O read requests, is read out to the memory cache pre-read, returns and the I O read requests Corresponding data.
The detailed process of step S830 to step S860 are referred to the step S110 to S140 in first embodiment, here It repeats no more.
The reading accelerated method based on IO scene Recognition that the 8th embodiment of the application provides records the associated IO of file destination Information, the IO that analysis obtains the file destination read scene behavior, obtain IO scene grouping set, I O read requests are received, by institute The scene and IO scene grouping set for stating I O read requests carry out similarity mode, determining and the I O read requests scene matchings IO scene is grouped library, and starting IO scene is pre-read, by IO corresponding with the IO scene of the scene matching of I O read requests grouping library Scene grouping feature library and the corresponding data in magnetic disk of the I O read requests read memory cache in advance, respond the I O read requests, right The memory cache pre-read is read out, and returns to data corresponding with the I O read requests.
9th embodiment
Referring to Fig. 9, Fig. 9 shows the reading accelerated method based on IO scene Recognition of the 9th embodiment of the application offer Flow diagram.It will be explained in detail below for process shown in Fig. 9, the reading acceleration side based on IO scene Recognition Method can specifically include following steps:
Step S901: the record associated IO information of file destination.
The detailed process of step S901 is referred to the step S810 in the 8th embodiment, and which is not described herein again.
Step S902: the analysis window t is determined by isochronous surface method；
The analysis window t is the IO speed and access duration empirical value of displaying, t=t0+ dT*N, N times incremented by successively Incremental time window, t0It is the initial analysis moment, dT is incremental time window.Empirically, dT can be with Initialize installation for 1, N 2 can be initialized as, so that the data volume that all IO are related to is substantially within 128MB.It can also be according in storage configuration Size is deposited, Initialize installation is adjusted.
Step S903: in the analysis window t, effective IO scene grouping set is calculated.
In the analysis window t, effective IO scene grouping set is calculated, i.e., the IO scene is discontinuously divided Analysis, finds the knot for meeting the IO scene grouping set i.e. IO scene grouping set of validity of the condition of continuity namely continuously analyzing Fruit, and save the IO scene grouping set for meeting the condition of continuity.
Step S904: results set fR { sR1, sR2, sR3... } is recorded in the last continuous analysis result.
Meet that the grouping set of the condition of continuity continuously analyzes obtained in step S903 as a result, by described the last time Continuous analysis result is recorded in results set fR { sR1, sR2, sR3... }.
Step S905: continuing since the discontinuous I O access moment, into next analysis window, until analysis is tied Beam.
Effective IO scene grouping set, which is calculated, in step S903 judges whether IO scene grouping set meets continuously Condition rejects the scene for meeting discontinuity condition using judging whether IO scene is discontinuous during concrete analysis, Need to record the discontinuous I O access moment at this time, then, continue since the discontinuous I O access moment, entrance is next A analysis window continues whether effectively to judge IO scene, until analysis terminates.
Step S906: retain the M scene grouping feature library IO to the IO grouping set.
M IO scene grouping feature library for meeting validity in the analysis window is remained into IO grouping set, shape At history I/O data library.It is to have Continuity Analysis spy comprising M IO scene grouping feature library volume IO scene grouping set The IO sequence grouping set of sign.
Step S907: I O read requests are received.
Step S908: the scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining and institute State the corresponding IO scene grouping feature library in IO scene grouping library of the scene matching of I O read requests.
Step S909: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
Step S910: responding the I O read requests, is read out to the memory cache pre-read, returns and the I O read requests Corresponding data.
The detailed process of step S907 to step S910 are referred to the step S830 to S860 in the 8th embodiment, here It repeats no more.
The reading accelerated method based on IO scene Recognition that the 9th embodiment of the application provides records the associated IO of file destination Information determines the analysis window t by isochronous surface method, in the analysis window t, calculates effective IO scene set of packets It closes, the last continuous analysis result is recorded results set fR { sR1, sR2, sR3... }, is continued from discontinuous IO The moment is accessed, into next analysis window, is terminated until analyzing, retains the M scene grouping feature library IO to the IO Grouping set receives I O read requests, and the scene of the I O read requests and IO scene grouping set are carried out similarity mode, determined IO scene grouping feature library corresponding with the IO scene of the scene matching of I O read requests grouping library, starting IO scene are pre-read, The corresponding IO scene grouping feature library in library and the I O read requests will be grouped with the IO scene of the scene matching of the I O read requests Corresponding data in magnetic disk reads memory cache in advance, responds the I O read requests, is read out to the memory cache pre-read, return with The corresponding data of the I O read requests.
Tenth embodiment
Referring to Fig. 10, Figure 10 shows the reading accelerated method based on IO scene Recognition of the tenth embodiment of the application offer Flow diagram.It will be explained in detail below for process shown in Fig. 10, the reading based on IO scene Recognition adds Fast method can specifically include following steps:
Step S1001: the record associated IO information of file destination.
Step S1002: the analysis window t is determined by isochronous surface method.
The detailed process of step S1001 to step S1002 are referred to the step S901 in the 9th embodiment to step S902, which is not described herein again.
Step S1003: the original record set rR in the analysis window t is extracted.
The original record set rR includes the information of I O access, it is therefore desirable to be extracted original in the analysis window t Set of records ends rR.
Step S1004: I O access feature is extracted for the original record set rR.
The process that I O access feature is extracted for the original record set rR is to rR pairs of the original record set The process that the access record sR answered is discontinuously judged.
Step S1005: results set fR { sR1, sR2, sR3... } is recorded in the last continuous analysis result；
Step S1006: continuing since the discontinuous I O access moment, into next analysis window, until analysis is tied Beam.
Step S1007: retain M IO feature scene and be grouped library to the IO grouping set.
Step S1008: I O read requests are received.
Step S1009: the scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining and institute State the corresponding IO scene grouping feature library in IO scene grouping library of the scene matching of I O read requests.
Step S1010: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
Step S1011: responding the I O read requests, is read out to the memory cache pre-read, returns to the I O read requests Corresponding data.
The detailed process of step S1005 to step S1011 are referred to the step S904 in the 9th embodiment to step S910, which is not described herein again.
The reading accelerated method based on IO scene Recognition that the tenth embodiment of the application provides records the associated IO of file destination Information determines the analysis window t by isochronous surface method, calculates effective IO scene grouping set, extract the analysis Original record set rR in window t extracts I O access feature for the original record set rR, will be the last continuous Analysis result is recorded results set fR { sR1, sR2, sR3... }, continues since the discontinuous I O access moment, under One analysis window, until analysis terminates, M IO feature scene of reservation is grouped library to the IO grouping set, receives IO reading and asks It asks, the scene of the I O read requests and IO scene grouping set is subjected to similarity mode, the determining field with the I O read requests The corresponding IO scene grouping feature library in scape matched IO scene grouping library, starting IO scene are pre-read, by with the I O read requests The IO scene grouping corresponding IO scene grouping feature library in library of scene matching and the corresponding data in magnetic disk of the I O read requests are pre-read To memory cache, the I O read requests are responded, the memory cache pre-read is read out, are returned corresponding with the I O read requests Data.
11st embodiment
Figure 11 is please referred to, Figure 11 shows the reading acceleration side based on IO scene Recognition of the 11st embodiment of the application offer The flow diagram of method.It will be explained in detail below for process shown in Figure 11, the reading based on IO scene Recognition Accelerated method can specifically include following steps:
Step S1101: the record associated IO information of file destination.
Step S1102: the analysis window t is determined by isochronous surface method.
Step S1103: the original record set rR in the analysis window t is extracted.
The detailed process of step S1101 to step S1103 are referred to the step S1001 in the tenth embodiment to step S1003, which is not described herein again.
Step S1104: according to the reading offset in the IO information of the original record set rR, all access of descending sort Record sR.
According to the reading offset in the IO information of the original record set rR, all access of descending sort record sR, drop The purpose of sequence sequence is easy for discontinuously analyzing.
Step S1105: when any I O access is discontinuous, terminating analysis window analysis, and writes down discontinuous at this time The I O access moment.
The I O access discontinuity condition is within 1 second time, and I O access number is lower than 100 times；Or for the visit It asks record sR, calculates the time difference of adjacent I O access twice, it is poor that the IO time difference meets preset time.
When any I O access is discontinuous, i.e., when any I O access meets discontinuity condition, terminate analysis window Mouth analysis, writes down the discontinuous I O access moment at this time.
Step S1106: IO offset is read in grouping as unit of the CFL of constant offset interval.
The constant offset interval CFL is an empirical value, can be set to 1MB.I O read requests are usually with 64KB for one A reading unit, 1MB are the integral multiples of 64KB, while being also corresponding bottom disk storage data block file size, convenient primary It is sequentially read from disk.
Step S1107: the analysis result that there is record continuous IO to deviate to IO scene grouping set sR sR1, sR2, sR3,...}。
The IO scene grouped record continuously analyzed will be met to IO scene grouping set sR { sR1, sR2, sR3 ... }.
Step S1108: the corresponding indexing parameter in each IO scene grouping feature library is calculated.
There is an indexing parameter in each IO scene grouping feature library, and the indexing parameter is to quickly read institute State IO scene grouping feature library, the IO scene grouping set in IO scene grouping feature library, that is, step S1107.The index Parameter is i.e. in the corresponding first IO offset of each analysis time window.
Step S1109: results set fR { sR1, sR2, sR3... } is recorded in the last continuous analysis result.
Step S1110: continuing since the discontinuous I O access moment, into next analysis window, until analysis is tied Beam.
Step S1111: retain M IO feature scene and be grouped library to the IO grouping set.
Step S1112: I O read requests are received.
Step S1113: the scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining and institute State the corresponding IO scene grouping feature library in IO scene grouping library of the scene matching of I O read requests.
Step S1114: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
Step S1115: responding the I O read requests, is read out to the memory cache pre-read, returns and asks with IO reading Seek corresponding data.
The detailed process of step S1109 to step S1115 are referred to the step S1005 in the tenth embodiment to step S1011, which is not described herein again.
The reading accelerated method based on IO scene Recognition that the 11st embodiment of the application provides, record file destination are associated IO information determines the analysis window t by isochronous surface method, extracts the original record set rR in the analysis window t, According to the reading offset in the IO information of the original record set rR, all access of descending sort record sR, when any described When I O access is discontinuous, terminate analysis window analysis, and write down the discontinuous I O access moment at this time, with constant offset interval CFL reads IO for unit grouping and deviates, analysis result of the record with continuous IO offset to IO scene grouping set sR sR1, sR2, SR3 ... }, the corresponding indexing parameter in each IO scene grouping feature library is calculated, by the last continuous analysis result It is recorded results set fR { sR1, sR2, sR3... }, continues since the discontinuous I O access moment, into next analysis Window, until analysis terminates, M IO feature scene of reservation is grouped library to the IO grouping set, receives I O read requests, will be described The scene and IO scene grouping set of I O read requests carry out similarity mode, the determining IO with the scene matching of the I O read requests Scene is grouped the corresponding IO scene grouping feature library in library, and starting IO scene is pre-read, by the scene matching with the I O read requests The IO scene grouping corresponding IO scene grouping feature library in library and the corresponding data in magnetic disk of the I O read requests read memory cache in advance, The I O read requests are responded, the memory cache pre-read is read out, return to data corresponding with the I O read requests.
12nd embodiment
Figure 12 is please referred to, Figure 12 shows the reading accelerated method based on IO scene Recognition of the 9th embodiment of the application offer Flow diagram.It will be explained in detail below for process shown in Figure 12, the reading based on IO scene Recognition adds Fast method can specifically include following steps:
Step S1201: the record associated IO information of file destination.
Step S1202: the analysis window t is determined by isochronous surface method.
Step S1203: in the analysis window t, effective IO scene grouping set is calculated.
Step S1204: results set fR { sR1, sR2, sR3... } is recorded in the last continuous analysis result.
The detailed process of step S1201 to step S1204 are referred to the step S901 to S904 in the 9th embodiment, this In repeat no more.
Step S1205: since the discontinuous I O access moment, again using N times of incremental window as analysis window.
Since the I O access moment discontinuous recorded in step S1203, analysis window is reset, i.e., again Analysis window is set with N times of incremental window.
Step S1206: in analysis window, effective IO scene grouping set is computed repeatedly and by the last continuous point Results set is recorded in analysis result, until analysis terminates.
In the analysis window reset in step S1205, repeat to calculate the IO grouping set in analysis window Effective grouping set, and results set is recorded by the grouping set continuously analyzed is met, until analysis terminates.
Step S1207: retain the M scene grouping feature library IO to the IO grouping set.
Step S1208: I O read requests are received.
Step S1209: the scene of the I O read requests and IO scene grouping set are subjected to similarity mode, determining and institute State the corresponding IO scene grouping feature library in IO scene grouping library of the scene matching of I O read requests.
Step S1210: starting IO scene is pre-read, will be corresponding with the IO scene of the scene matching of I O read requests grouping library The scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests read memory cache in advance.
Step S1211: responding the I O read requests, is read out to the memory cache pre-read, returns to the I O read requests Corresponding data.
The detailed process of step S1207 to step S1211 are referred to the step S906 to S910 in first embodiment, this In repeat no more.
The reading accelerated method based on IO scene Recognition that the 12nd embodiment of the application provides, record file destination are associated IO information determines the analysis window t by isochronous surface method, in the analysis window t, calculates effective IO scene grouping Results set fR { sR1, sR2, sR3... } is recorded in the last continuous analysis result by set, from described discontinuous The I O access moment, in analysis window, computes repeatedly effective IO scene again using N times of incremental window as analysis window Grouping set and results set fR { sR1, sR2, sR3 ... } is recorded in the last continuous analysis result, until analysis Terminate, retains the M scene grouping feature library IO to the IO grouping set, I O read requests are received, by the field of the I O read requests Scape and IO scene grouping set carry out similarity mode, and the determining IO scene with the scene matching of the I O read requests is grouped library pair The IO scene grouping feature library answered, starting IO scene are pre-read, and the IO scene of the scene matching with the I O read requests is grouped library Corresponding IO scene grouping feature library and the corresponding data in magnetic disk of the I O read requests read memory cache in advance, respond the IO and read Request, is read out the memory cache pre-read, returns to data corresponding with the I O read requests.
13rd embodiment
The 13rd embodiment of the application provides a kind of reading based on IO scene Recognition and accelerates 1300, and referring to Figure 13, it should The managing device 1300 of home equipment includes: receiving module 1310, processing module 1320, pre- read through model 1330 and respond module 1340.Wherein, the receiving module 1310 is for receiving I O read requests；
The processing module is used to the scene of the I O read requests and IO scene grouping set carrying out similarity mode, really The fixed IO scene with the scene matching of the I O read requests is grouped library；
The pre- read through model is pre-read for starting IO scene, and the IO scene of the scene matching with the I O read requests is grouped Library and the corresponding data in magnetic disk of the I O read requests read memory cache in advance；
The corresponding module is read out the memory cache pre-read, returns to the IO for responding the I O read requests The corresponding data of read request.
In conclusion compared with the existing technology, the reading accelerated method provided by the present application based on IO scene Recognition, device, And the scene of the I O read requests and IO scene grouping set are carried out similarity by receiving I O read requests by electronic equipment Match, determining IO scene grouping feature library corresponding with the IO scene of the scene matching of I O read requests grouping library starts IO Scape is pre-read, and will be grouped the corresponding IO scene grouping feature library in library and the IO with the IO scene of the scene matching of the I O read requests The corresponding data in magnetic disk of read request reads memory cache in advance, responds the I O read requests, is read out to the memory cache pre-read, The corresponding data of the I O read requests are returned, IO is realized and reads to accelerate.
It should be noted that all the embodiments in this specification are described in a progressive manner, each embodiment weight Point explanation is all differences from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For device class embodiment, since it is basically similar to the method embodiment, so being described relatively simple, related place ginseng See the part explanation of embodiment of the method.For arbitrary processing mode described in embodiment of the method, in device reality Apply in example can no longer be repeated in Installation practice by corresponding processing modules implement one by one.
Figure 14 is please referred to, the processing method of the reading acceleration based on above-mentioned IO scene Recognition, device, the embodiment of the present application is also A kind of terminal device 100 is provided, the terminal device 100 usually may include one or more (one is only shown in figure) processing Device 102, memory 104, RF (Radio Frequency, radio frequency) module 106, power module 108.Ordinary skill people Member does not cause to limit it is appreciated that structure shown in Fig. 10 only signal to the structure of the terminal 100.For example, described Terminal device 100 may also include than shown in Figure 10 more perhaps less component or with pair different from shown in Figure 10 It answers.
It will appreciated by the skilled person that every other component belongs to for the processor 102 It is coupled between peripheral hardware, the processor 102 and these peripheral hardwares by multiple Peripheral Interfaces 110.The Peripheral Interface 110 can Based on following standard implementation: Universal Asynchronous Receive/sending device (Universal Asynchronous Receiver/ Transmitter, UART), universal input/output (General Purpose Input Output, GPIO), serial peripheral connect Mouthful (Serial Peripheral Interface, SPI), internal integrated circuit (Inter-Integrated Circuit, I2C), but it is not limited to above-mentioned standard.In some instances, the Peripheral Interface 110 can only include bus；In other examples In, the Peripheral Interface 110 may also include other elements, such as one or more controller.In addition, these controllers can be with It detaches, and is integrated in the processor 102 or in corresponding peripheral hardware from the Peripheral Interface 110.
The memory 104 can be used for storing software program and module, and the processor 102 is stored in institute by operation The software program and module in memory 104 are stated, thereby executing various function application and data processing.The memory 104 may include high speed random access memory, may also include nonvolatile memory, and such as one or more magnetic storage device dodges It deposits or other non-volatile solid state memories.In some instances, the memory 104 can further comprise relative to institute The remotely located memory of processor 102 is stated, these remote memories can pass through network connection to the terminal device 100. The example of above-mentioned network includes but is not limited to internet, intranet, local area network, mobile radio communication and combinations thereof.
The RF module 106 is used to receive and transmit electromagnetic wave, realizes the mutual conversion of electromagnetic wave and electric signal, thus It is communicated with communication network or other equipment.The RF module 106 may include various existing for executing these functions Circuit element, for example, antenna, RF transceiver, digital signal processor, encryption/deciphering chip, subscriber identity module (SIM) card, memory etc..The RF module 106 can be carried out with various networks such as internet, intranet, wireless network Communication is communicated by wireless network and other equipment.Above-mentioned wireless network may include cellular telephone networks, wireless Local area network or Metropolitan Area Network (MAN).Various communication standards, agreement and technology can be used in above-mentioned wireless network, including but not limited to Global system for mobile communications (Global System for Mobile Communication, GSM), enhanced mobile communication skill Art (Enhanced Data GSM Environment, EDGE), Wideband CDMA Technology (wideband code Division multiple access, W-CDMA), Code Division Multiple Access (Code division access, CDMA), time-division Multiple access technology (time division multiple access, TDMA), adopting wireless fidelity technology (Wireless, Fidelity, WiFi) (such as American Institute of Electrical and Electronics Engineers's standard IEEE 802.10A, IEEE 802.11b, IEEE802.11g and/or IEEE 802.11n), the networking telephone (Voice over internet protocal, VoIP), worldwide interoperability for microwave accesses (Worldwide Interoperability for Microwave Access, Wi-Max), other be used for mail, Instant Messenger The agreement and any other suitable communications protocol of news and short message, or even may include that those are not developed currently yet Agreement.
The power module 108 is used to provide power supply to the processor 102 and other each components.Specifically, The power module 108 may include power-supply management system, one or more power supply (such as battery or alternating current), charging circuit, Power-fail detection circuit, inverter, indicator of the power supply status and any other life with electric power in the terminal device 100 At, manage and be distributed relevant component.
In the description of this specification, reference term " one embodiment ", " some embodiments ", " example ", " specifically show The description of example " or " some examples " etc. means specific features, structure, material or spy described in conjunction with this embodiment or example Point is contained at least one embodiment or example of the application.In the present specification, schematic expression of the above terms are not It must be directed to identical embodiment or example.Moreover, particular features, structures, materials, or characteristics described can be in office It can be combined in any suitable manner in one or more embodiment or examples.In addition, without conflicting with each other, the skill of this field Art personnel can tie the feature of different embodiments or examples described in this specification and different embodiments or examples It closes and combines.
In addition, term " first ", " second " are used for descriptive purposes only and cannot be understood as indicating or suggesting relative importance Or implicitly indicate the quantity of indicated technical characteristic.Define " first " as a result, the feature of " second " can be expressed or Implicitly include at least one this feature.In the description of the present application, the meaning of " plurality " is at least two, such as two, three It is a etc., unless otherwise specifically defined.
Any process described otherwise above or method description are construed as in flow chart or herein, and expression includes It is one or more for realizing specific logical function or process the step of executable instruction code module, segment or portion Point, and the range of the preferred embodiment of the application includes other realization, wherein can not press shown or discussed suitable Sequence, including according to related function by it is basic simultaneously in the way of or in the opposite order, to execute function, this should be by the application Embodiment person of ordinary skill in the field understood.
Expression or logic and/or step described otherwise above herein in flow charts, for example, being considered use In the order list for the executable instruction for realizing logic function, may be embodied in any computer-readable medium, for Instruction execution system, device or equipment (such as computer based system, including the system of processor or other can be held from instruction The instruction fetch of row system, device or equipment and the system executed instruction) it uses, or combine these instruction execution systems, device or set It is standby and use.For the purpose of this specification, " computer-readable medium ", which can be, any may include, stores, communicates, propagates or pass Defeated program is for instruction execution system, device or equipment or the dress used in conjunction with these instruction execution systems, device or equipment It sets.The more specific example (non-exhaustive list) of computer-readable medium include the following: there is the electricity of one or more wirings Interconnecting piece (mobile terminal), portable computer diskette box (magnetic device), random access memory (RAM), read-only memory (ROM), erasable edit read-only storage (EPROM or flash memory), fiber device and portable optic disk is read-only deposits Reservoir (CDROM).In addition, computer-readable medium can even is that the paper that can print described program on it or other are suitable Medium, because can then be edited, be interpreted or when necessary with it for example by carrying out optical scanner to paper or other media His suitable method is handled electronically to obtain described program, is then stored in computer storage.
It should be appreciated that each section of the application can be realized with hardware, software, firmware or their combination.Above-mentioned In embodiment, software that multiple steps or method can be executed in memory and by suitable instruction execution system with storage Or firmware is realized.It, and in another embodiment, can be under well known in the art for example, if realized with hardware Any one of column technology or their combination are realized: having a logic gates for realizing logic function to data-signal Discrete logic, with suitable combinational logic gate circuit specific integrated circuit, programmable gate array (PGA), scene Programmable gate array (FPGA) etc..
Those skilled in the art are understood that realize all or part of step that above-described embodiment method carries It suddenly is that relevant hardware can be instructed to complete by program, the program can store in a kind of computer-readable storage medium In matter, which when being executed, includes the steps that one or a combination set of embodiment of the method.In addition, in each embodiment of the application In each functional unit can integrate in a processing module, be also possible to each unit and physically exist alone, can also two A or more than two units are integrated in a module.Above-mentioned integrated module both can take the form of hardware realization, can also It is realized in the form of using software function module.If the integrated module realized in the form of software function module and as Independent product when selling or using, also can store in a computer readable storage medium.
Storage medium mentioned above can be read-only memory, disk or CD etc..Although having been shown and retouching above Embodiments herein is stated, it is to be understood that above-described embodiment is exemplary, and should not be understood as the limit to the application System, those skilled in the art can be changed above-described embodiment, modify, replace and become within the scope of application Type.
Finally, it should be noted that above embodiments are only to illustrate the technical solution of the application, rather than its limitations；Although The application is described in detail with reference to the foregoing embodiments, those skilled in the art are when understanding: it still can be with It modifies the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；And These are modified or replaceed, do not drive corresponding technical solution essence be detached from each embodiment technical solution of the application spirit and Range.

Claims (19)

1. a kind of reading accelerated method based on IO scene Recognition, which is characterized in that be applied to distributed memory system, the method Include:
Receive I O read requests；
The scene of the I O read requests and IO scene grouping set are subjected to similarity mode, the determining field with the I O read requests The corresponding IO scene grouping feature library in scape matched IO scene grouping library；
Starting IO scene is pre-read, and IO scene corresponding with the IO scene of the scene matching of I O read requests grouping library is grouped special Sign library and the corresponding data in magnetic disk of the I O read requests read memory cache in advance；
The I O read requests are responded, the memory cache pre-read is read out, return to the corresponding data of the I O read requests.
2. the method as described in claim 1, which is characterized in that the scene by the I O read requests and IO feature scene point Group set carries out similarity mode, determining IO scene point corresponding with the IO scene of the scene matching of I O read requests grouping library Group feature database, comprising:
Initialization time window t1；
Count the time window t1The scene characteristic of the interior I O read requests；
The scene characteristic of the I O read requests is matched with the IO scene grouping library in IO scene grouping set；
Obtain IO scene grouping feature library corresponding with the IO scene of the scene matching of I O read requests grouping library.
3. method according to claim 2, which is characterized in that the time window t1It is initialized as the N% of analysis window t.
4. method according to claim 2, which is characterized in that the scene characteristic by the I O read requests and IO scene point IO scene grouping library in group set carries out matching corresponding IO scene grouping feature library, comprising:
The IO scene grouping set is scanned, by the IO in the scene characteristic of the I O read requests and the IO scene grouping set Scene is grouped library and carries out similarity mode；
Determining IO scene grouping feature library corresponding with the IO scene of the scene characteristic successful match of I O read requests grouping library.
5. as claimed in claim 4, which is characterized in that the scanning IO scene grouping set, by the I O read requests IO scene grouping library in scene characteristic and the IO scene grouping set carries out similarity mode, comprising:
Set number tc, count the average offset value a of the IO scene grouping set0；
Based on first IO offset, in the continuous number tcIt is interior, the I O read requests are obtained in the time window t1 Interior offset average value a1, calculate a1With the average offset value a0The difference subtracted each other, then divided by the average offset value a0Quotient Value, the quotient and 0 size；
If the quotient is equal to 0, the scene and IO scene grouping storehouse matching success of the I O read requests are determined, then by institute The corresponding deviant of first IO of I/O Request is stated as indexing parameter；
If the quotient is not equal to 0, continue in next continuous number tcInterior calculating, until can not find matching offset.
6. the method as described in claim 1, which is characterized in that the starting IO scene is pre-read, by with the I O read requests The IO scene grouping corresponding IO scene grouping feature library in library of scene matching and the corresponding data in magnetic disk of the I O read requests are pre-read To memory cache, comprising:
Starting IO scene is pre-read；
The corresponding IO scene grouping feature library in library will be grouped with the IO scene of the scene matching of the I O read requests and the IO is read Corresponding data in magnetic disk is requested to be saved in memory cache；
Setting time K, when in the setting time K to the memory cache without access, discharge the memory cache.
7. the method as described in claim 1, which is characterized in that I O read requests are received, by the scene and IO of the I O read requests Feature scene grouping set carries out similarity mode, determining corresponding with the IO scene of the scene matching of I O read requests grouping library IO scene grouping feature library, starting IO scene pre-reads, and the IO scene of the scene matching with the I O read requests is grouped library pair Memory cache is read in the scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests answered in advance, is responded the IO reading and is asked It asks, the memory cache pre-read is read out, return to the corresponding data of the I O read requests, comprising:
Access includes the memory cache with the IO scene of the scene matching of I O read requests grouping library；
If being not present in the memory cache of the access and being grouped library with the IO scene of the scene matching of the I O read requests, after Continue to memory node and initiates read requests；
If existing in the memory cache of the access and being grouped library with the IO scene of the scene matching of the I O read requests, return The corresponding data of the I O read requests.
8. the method as described in claim 1, which is characterized in that before the reception I O read requests further include:
Record the associated IO information of file destination；
The IO that analysis obtains the file destination reads scene behavior, obtains IO scene grouping set.
9. method according to claim 8, which is characterized in that the associated IO information of record file destination includes target text Part ID, the time, offset, data block fingerprint and data block location are read.
10. method according to claim 8, which is characterized in that described analyze obtains the IO reading scene row of the file destination To obtain IO scene grouping set, comprising:
The analysis window t is determined by isochronous surface method；
In the analysis window t, effective IO scene grouping set is calculated；
Results set fR { sR1, sR2, sR3... } is recorded in the last continuous analysis result；
Continue since the discontinuous I O access moment, into next analysis window, until analysis terminates；
Retain the M scene grouping feature library IO to the IO grouping set.
11. method as claimed in claim 10, which is characterized in that described to determine the analysis window by isochronous surface method T, comprising:
For the analysis window t, t=t0+ dT*N, N times of incremental time window incremented by successively, t0It is the initial analysis moment, dT is Incremental time window.
12. method as claimed in claim 10, which is characterized in that it is described in the analysis window t, calculate effective IO scene Grouping set, comprising:
Extract the original record set rR in the analysis window t；
I O access feature is extracted for the original record set rR.
13. method as claimed in claim 12, which is characterized in that described to extract I O access for the original record set rR Feature, comprising:
According to the reading offset in the IO information of the original record set rR, all access of descending sort record sR；
When any I O access is discontinuous, when terminating the analysis of this analysis window, and writing down discontinuous I O access at this time It carves；
IO offset is read in grouping as unit of the CFL of constant offset interval；
Recording has the analysis result of continuous IO offset to IO scene grouping set sR { sR1, sR2, sR3 ... }；
Calculate the corresponding indexing parameter in each IO feature scene grouping library.
14. method as claimed in claim 13, which is characterized in that the I O access discontinuously includes:
Within 1 second time, I O access number is lower than 100 times；Or
SR is recorded for the access, calculates the time difference of adjacent I O access twice, it is poor that the IO time difference meets preset time.
15. method as claimed in claim 13, which is characterized in that calculate the corresponding rope in each IO feature scene grouping library Draw, comprising:
It is corresponding index value in the corresponding first IO offset of each analysis time window.
16. method as claimed in claim 10, which is characterized in that it is described to continue since the discontinuous I O access moment, into Enter next analysis window, until analysis terminates, comprising:
Since the discontinuous I O access moment, again using N times of incremental window as analysis window；
In analysis window, computes repeatedly effective IO scene grouping set and knot is recorded in the last continuous analysis result Fruit set, until analysis terminates.
17. method as claimed in claim 10, which is characterized in that M scene grouping feature library IO of the reservation to the IO Grouping set, comprising:
The IO grouping set with M IO scene grouping feature library is the IO sequence set of packets for having Continuity Analysis feature It closes.
18. a kind of reading accelerator based on IO scene Recognition, which is characterized in that described device includes: receiving module, processing mould Block, pre- read through model and respond module, wherein
The receiving module is for receiving I O read requests；
The processing module is used to the scene of the I O read requests and IO scene grouping set carrying out similarity mode, determine with The corresponding IO scene grouping feature library in IO scene grouping library of the scene matching of the I O read requests；
The pre- read through model is pre-read for starting IO scene, and the IO scene of the scene matching with the I O read requests is grouped library pair Memory cache is read in the scene grouping feature library IO and the corresponding data in magnetic disk of the I O read requests answered in advance；
The corresponding module is read out the memory cache pre-read, returns to the IO reading and ask for responding the I O read requests Seek corresponding data.
19. a kind of electronic equipment, which is characterized in that including memory and processor, the memory is couple to the processing Device, the memory store instruction, the processor executes such as claim 1- when executed by the processor 17 described in any item methods.
CN201810866060.XA 2018-06-05 2018-08-01 Read acceleration method and device based on IO scene recognition and electronic equipment Active CN109343789B (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201810568685 2018-06-05
CN2018105686858 2018-06-05

Publications (2)

Publication Number Publication Date
CN109343789A true CN109343789A (en) 2019-02-15
CN109343789B CN109343789B (en) 2021-08-17

Family

ID=65291499

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810866060.XA Active CN109343789B (en) 2018-06-05 2018-08-01 Read acceleration method and device based on IO scene recognition and electronic equipment

Country Status (1)

Country Link
CN (1) CN109343789B (en)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111552740A (en) * 2020-04-28 2020-08-18 腾讯科技（深圳）有限公司 Data processing method and device
CN112835853A (en) * 2020-12-31 2021-05-25 北京聚云科技有限公司 Data processing type determination method and device
CN114115719A (en) * 2021-08-24 2022-03-01 深圳市木浪云科技有限公司 IO batch processing method and device based on IO mode identification and storage medium
CN114461588A (en) * 2021-08-20 2022-05-10 荣耀终端有限公司 Method for adjusting pre-reading window and electronic equipment

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102340416A (en) * 2011-07-08 2012-02-01 东软集团股份有限公司 Time slice-based method and device for event statistics
CN103677900A (en) * 2013-11-15 2014-03-26 北京奇虎科技有限公司 Method and device for accelerating starting of computer device
US8713260B2 (en) * 2010-04-02 2014-04-29 Intel Corporation Adaptive block pre-fetching method and system
CN104572205A (en) * 2015-01-12 2015-04-29 安一恒通（北京）科技有限公司 Method and device for software acceleration
CN107340978A (en) * 2017-07-18 2017-11-10 郑州云海信息技术有限公司 One kind storage pre-head method, device and storage system

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8713260B2 (en) * 2010-04-02 2014-04-29 Intel Corporation Adaptive block pre-fetching method and system
CN102340416A (en) * 2011-07-08 2012-02-01 东软集团股份有限公司 Time slice-based method and device for event statistics
CN103677900A (en) * 2013-11-15 2014-03-26 北京奇虎科技有限公司 Method and device for accelerating starting of computer device
CN104572205A (en) * 2015-01-12 2015-04-29 安一恒通（北京）科技有限公司 Method and device for software acceleration
CN107340978A (en) * 2017-07-18 2017-11-10 郑州云海信息技术有限公司 One kind storage pre-head method, device and storage system

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111552740A (en) * 2020-04-28 2020-08-18 腾讯科技（深圳）有限公司 Data processing method and device
CN111552740B (en) * 2020-04-28 2023-09-12 腾讯科技（深圳）有限公司 Data processing method and device
CN112835853A (en) * 2020-12-31 2021-05-25 北京聚云科技有限公司 Data processing type determination method and device
CN112835853B (en) * 2020-12-31 2024-03-22 北京聚云科技有限公司 Data processing type determining method and device
CN114461588A (en) * 2021-08-20 2022-05-10 荣耀终端有限公司 Method for adjusting pre-reading window and electronic equipment
CN114461588B (en) * 2021-08-20 2023-01-24 荣耀终端有限公司 Method for adjusting pre-reading window and electronic equipment
CN114115719A (en) * 2021-08-24 2022-03-01 深圳市木浪云科技有限公司 IO batch processing method and device based on IO mode identification and storage medium
CN114115719B (en) * 2021-08-24 2022-10-18 深圳市木浪云科技有限公司 IO batch processing method and device based on IO mode identification and storage medium

Also Published As

Publication number Publication date
CN109343789B (en) 2021-08-17

Similar Documents

Publication Publication Date Title
CN109343789A (en) A kind of reading accelerated method, device and electronic equipment based on IO scene Recognition
CN105847447B (en) Message pushing method and device
CN109669943A (en) Date storage method, data query method and device
CN107798046B (en) Client identity information processing method, device, storage medium and computer equipment
CN110505141A (en) Processing method, device, readable medium and the electronic equipment of instant communication information
CN107294974A (en) The method and apparatus for recognizing target clique
CN111488529B (en) Information processing method, information processing apparatus, server, and storage medium
CN113673260A (en) Model processing method, device, storage medium and processor
CN106850149A (en) A kind of data transmission method, device, equipment and storage medium
CN107872533B (en) Information-pushing method, device, server and storage medium
WO2022141498A1 (en) Charging method, apparatus and system
CN106488055A (en) Calling list rearrangement method, back end equipment and routing node device
CN109104723A (en) A kind of network name display methods and relevant device
CN115798513A (en) Talent expression management method, system and computer readable storage medium
CN105677478A (en) Method and apparatus for resources management
CN110505143A (en) It is a kind of for sending the method and apparatus of target video
CN108989902A (en) A kind of processing method, device, terminal and the storage medium of barrage message
CN109299454A (en) Abstraction generating method and device, storage medium and electric terminal based on chat log
CN108021639A (en) Information-pushing method, device, server and storage medium
WO2021248316A1 (en) Event processing method and apparatus, electronic device, and storage medium
CN108011941B (en) Content delivery method, device, server and storage medium
CN109146617A (en) A kind of business approval method, server, client and storage medium
CN108156011B (en) Method and equipment for clustering wireless access points
CN106599121B (en) Data processing method and system for Internet of things
CN112566054B (en) Method and system for optimizing message interaction process

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for 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: 20210918

Address after: 518103 514, building 1, hengtaiyu building, Tangwei community, Fenghuang street, Guangming District, Shenzhen, Guangdong

Patentee after: Shenzhen mulangyun Technology Co.,Ltd.

Address before: 518000 307, block a, Languang science and Technology Park, No. 7, songpingshan new West Road, North District, Xili street, Nanshan District, Shenzhen, Guangdong

Patentee before: SHENZHEN MULANG CLOUD DATA Co.,Ltd.

TR01 Transfer of patent right