CN102023891A - Concurrent garbage collector frame based on Java virtual machine - Google Patents
Concurrent garbage collector frame based on Java virtual machine Download PDFInfo
- Publication number
- CN102023891A CN102023891A CN 201010595749 CN201010595749A CN102023891A CN 102023891 A CN102023891 A CN 102023891A CN 201010595749 CN201010595749 CN 201010595749 CN 201010595749 A CN201010595749 A CN 201010595749A CN 102023891 A CN102023891 A CN 102023891A
- Authority
- CN
- China
- Prior art keywords
- concurrent
- thread
- collection
- state
- algorithm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The invention belongs to the technical field of Java virtual machine garbage retriever, particularly relating to a concurrent garbage collector frame based on a Java virtual machine. The frame is divided into two modules of a concurrent algorithm frame and a dispatcher, wherein the concurrent algorithm frame is designed to mainly carry out variation analysis on the existing concurrent garbage collection algorithm to finally divide modules which are conviennt to be reused or can not be reused, and the division is realized by aiming at different characteristics; the dispatcher is mainly composed of a dispatching trigger module, a dispatching algorithm module and an exhaustion processing module and mainly has two functions on determining when to trigger GC (garbage collection) and to process the situation of system resource exhaustion. According to the invention, the platform transportability of the concurrent garbage collection technology is increased to solve the problem that different concurrent garbage collection technologies are experimented under the same platform, and the superiority or inferiority of different concurrent garbage collection technologies can be researched under a certain specific platform.
Description
Technical field
The invention belongs to Java Virtual Machine Garbage Collector technical field, be specifically related to a kind of concurrent Garbage Collector framework based on Java Virtual Machine.
Background technology
Be accompanied by the rise of high-level programming language, (Garbage Collection, GC) technology is being played the part of more and more important role aspect security of system and the resource management in refuse collection.
In a lot of programing systems, the obsolete memory headroom of release that the programming person need show in the programming process.In this design, programming person's error is easy to cause the internal memory of program to reveal or abnormal end, and this has influenced the stability and the reliability of system to a great extent.And garbage reclamation technique requires the memory headroom that the virtual machine execution is judged and liquidation procedures does not re-use.So garbage collection techniques can be avoided the generation of these situations effectively, it makes the management automation of internal memory, the programming person no longer needs obsolete internal memory is discharged, and garbage collector can be handled this heap a suitable time, and the space that has not re-used is discharged.The management that such design makes internal memory is modularization more, and has higher reliability.
When common garbage collection algorithm exhausted at the system heap internal memory, all application program threads of virtual machine meeting " time-out " were carried out refuse collection.Finish Deng refuse collection work, application program threads can be restarted continues its execution.This technology is called as the refuse collection of " stopping the world (stop-the-world) ".Amount of ram increases, number of objects increases along with the modern Application program is used, and " stopping the world " pause that refuse collection caused has had a strong impact on the response time of application program.
So in order to eliminate the problem that " stopping the world " refuse collection causes performance impact, a lot of research institutions and individual are in the technology of being devoted to realize to make garbage collector carry out refuse collection under the situation that application program threads is carried out, improve the efficient of application program with this, such technology is called as concurrent garbage collection techniques.But at present existing concurrent garbage collection techniques realizes all relatively disperseing, and realization is more loaded down with trivial details, also is a bigger problem for the portable relatively poor of different systems.
Summary of the invention
The objective of the invention is to realize the relatively problem of dispersion, propose the Unified frame of a concurrent garbage collector at existing concurrent garbage collection techniques.
The present invention has investigated the difference of different concurrent garbage collection algorithms, finding still has similarity significantly between the different concurrent garbage collection algorithms, only when realizing, concrete algorithm have bigger difference, such finding that possibility and the superiority of setting up Unified frame has been described.
The technical solution used in the present invention is: for Unified frame of concurrent garbage collector design, can realize and improve various concurrent garbage collection algorithm therein easily.And garbage collector itself is designed to multithreading (GC thread), improves the efficient of refuse collection with this.
The Unified frame of concurrent garbage collector is the special designed Unified frame of concurrent garbage collector that is.By studying the similarities and differences of different concurrent garbage collection algorithms, this framework provides general refuse collection data processing and functions such as merging, refuse collection.By this framework, the developer of concurrent garbage collector only need realize that its core algorithm just can realize a good concurrent garbage collector of work, greatly facilitates various concurrent garbage collection algorithm and realizes and improve.The Unified frame of concurrent garbage collector mainly is divided into concurrent algorithm frame and two modules of scheduler.
The concurrent garbage collector framework that the present invention proposes based on Java Virtual Machine, be divided into two parts of two modules of concurrent algorithm frame and scheduler, wherein, described concurrent algorithm frame, mainly carry out variance analysis in design by concurrent garbage collection algorithm to current existence, be convenient to reuse and the module that can not reuse in the final partitioning algorithm, and realize respectively at different characteristics; Again whole concurrent algorithm frame is divided into core algorithm realization, data parallel processing and merging, these three modules of concurrent collection in the existing framework; Described scheduler mainly by scheduling trigger, dispatching algorithm, exhaust and handle three modules and form; This scheduler mainly contains the effect of two aspects:
The first, when decision triggers GC;
The second, the situation of disposal system resource exhaustion; Exhaust the moment section of generation according to system resource, the situation of exhausting is divided into three kinds: it is in service that one occurs in application program, and the refuse collection thread is not activated; Occur in the tracking phase after the refuse collection thread starts; And the collection clean-up phase after occurring in the refuse collection thread and starting; Scheduler is handled respectively these three kinds of situations:
(1) for first kind of situation, only all threads are all stopped, do once the garbage collection process of " stopping the world ", allow program continue operation afterwards again;
(2) exhaust tracking phase after appearing at the refuse collection thread and starting when system resource, scheduler is cut off the thread of other application programs, allows the refuse collection thread that tracking phase is finished, and obtains the object trace information earlier; The garbage collection process of doing once " stopping the world " with this object trace information comes releasing memory afterwards;
(3) exhaust the collection clean-up phase that appears at after the refuse collection thread starts for system resource, scheduler is only cut off the thread of other application programs, allows the refuse collection thread finish collecting clean-up phase.
The maximum characteristics of concurrent garbage collector are exactly that GC thread and application program threads are moved simultaneously, and application program still be garbage collector all may be multithreading.We designed a GC sets of threads and application program threads group state shared variable guarantee between the thread synchronously.Fig. 1 has described the change that thread carries out this state variable in the concurrent garbage collector operational process.
If the situation that the taking into account system heap memory exhausts in advance just can be expressed as 6 conversions between the state to garbage collection process, the transformational relation between them as shown in Figure 2, respectively the expression act as:
1, state 1.: application program is normally moved, and the operation of at every turn applying for memory headroom all can be called the GC scheduler and be detected whether need to trigger concurrent garbage collection process, and 2. the words that need forward state to;
2, state 2.: application program threads is paused respectively, collects the root node on its execution stack, puts into the root node set, and opens the blocker of writing of current thread.After the root node set collection of last thread is finished, start the GC thread and begin to carry out tracing process, forward state to 3.;
3, state 3.: 4. the operation that tracing process and application program are concurrent after last GC thread is finished tracing process, forwards state to;
4, state 4.: tracing process is finished, and can trigger concurrent collection process when application program has new application internal memory operation, forwards state to 5.;
5, state 5.: 1. collection process and application program concurrent running forward state to after last GC thread is finished collection process;
6, state 6.: garbage collection process is in state 1., 3., 5. the time, exhaust if the system heap internal memory is employed program, will get the hang of and 6. handle (the system heap internal memory exhausts processing), forwards state subsequently to 1..
In implementation procedure, should consider Research on Interactive Problem between a plurality of application program threads and the Research on Interactive Problem between a plurality of GC thread.So we add the shared cogradient variable of some a plurality of application program threads and the shared cogradient variable of a plurality of GC thread is handled these problems, corresponding with it state conversion process such as Fig. 3.
Above-mentioned this state exchange is that Unified frame that we design provides is concurrent garbage collector design specially, as long as utilize above-mentioned state transformation, the programmer can realize most concurrent garbage collection algorithms.Based on this Unified frame, three kinds of concurrent garbage collection algorithms have been realized waiting, they are two kinds of different realizations of Mostly-Concurrent algorithm and Snapshot-at-the-Beginning algorithm, and this has illustrated that also Unified frame can realize different parallel garbage collection techniques easily.Here also these three kinds of parallel garbage collection techniques are simply introduced:
The Mostly-Concurrent basic idea is that " mark " done as much as possible with concurrent GC thread earlier works, and uses once " stopping the world " tracing process to finish remaining " mark " work then; And the main thought of Snapshot-at-the-Beginning algorithm is respectively the running stack of each application program threads of current operation to be done once " snapshot " (snapshot), obtain a complete root node set, the GC thread is from the tracing process of this root node set beginning execution concurrence then.In embodiment one joint, will be described in detail these three kinds of algorithms and data structure.
The invention has the beneficial effects as follows: 1) Unified frame that realizes concurrent garbage collector is provided, various algorithm can be implemented easily under a framework.2) realization of framework is not at concrete a certain platform, and such design has increased the platform portability of parallel garbage collection techniques.3) solve the problem of under identical platform, testing different parallel garbage collection techniques, so just can under a specific platform, study the superiority-inferiority of different parallel garbage collection techniques.4) under the framework that makes up at present, design and Implement three kinds of concurrent garbage collection algorithms, effectively utilized system resource, proved the practicality of this framework.5) by test shows, concurrent garbage collector can effectively be eliminated the service end application program because of pause that refuse collection caused; For client application, also can when reducing the dead time, effectively shorten the time of the execution of application program.
Description of drawings
Figure 1 shows that the synchronized relation of application program threads and GC thread in the concurrent garbage collector.
Figure 2 shows that the state exchange of concurrent refuse collection.
Figure 3 shows that the state conversion process in the operational process of concurrent refuse collection.
The heap space that Figure 4 shows that concurrent garbage collector exhausts processing.
Embodiment
The present invention mainly is divided into concurrent algorithm frame and two modules of scheduler according to above-mentioned concurrent refuse collection framework.Below just a detailed introduction is done in the concrete enforcement of these two modules in this framework.When introducing concurrent algorithm frame, also can three kinds of concurrent garbage collection algorithms realizing based on this framework be introduced.
Fig. 2 is the state transition graph of concurrent refuse collection, and the operation of concurrent refuse collection framework is exactly the conversion of these 6 states.Preceding 5 states belong to the category of concurrent algorithm frame, and last the 6th state is that the memory source of realizing in the scheduler exhausts processing.
1) concurrent algorithm frame.Concurrent algorithm frame is the core of this whole Unified frame, mainly carry out variance analysis in design by concurrent garbage collection algorithm to current existence, be convenient to reuse and the module that can not reuse in the final partitioning algorithm, and realize respectively at different characteristics.Again whole concurrent algorithm frame core algorithm realization, data parallel processing and merging, these three modules of concurrent collection have been divided in the existing framework.
Find that through investigation the processing of a big chunk can unify to realize in concurrent garbage collection techniques.The operation of in a large number data of system being carried out concurrent processing and merging is arranged in for example concurrent garbage collection techniques, and being different concurrent garbage collection techniques, the operation of this class all needs, so Unified frame has reduced the workload difficulty that realizes concurrent garbage collection algorithm to a great extent by abstract and extract the module that a data parallel that is applicable to all concurrent garbage collection techniques is handled and merged.On the core concept and performing step that only are embodied in specific algorithm of the difference maximum between the different concurrent garbage collection techniques.Briefly, data parallel processing in the concurrent algorithm frame and merging module and concurrent collection module are applicable to all concurrent garbage collection techniques, realize different concurrent collection methods as long as realize that its core algorithm is partly just passable under this framework.
A) core algorithm is realized
B) data parallel is handled
C) concurrent collection method
2) scheduler.Scheduler mainly by scheduling trigger, dispatching algorithm, exhaust and handle three modules and form.For scheduler, mainly contain the effect of two aspects:
The first, scheduler need determine when trigger GC.Because if refuse collection takes place just not make full use of system memory resource and cause the frequency of refuse collection too fast too early, can influence application program capacity like this; If took place evening, the system heap internal memory may be depleted before concurrent refuse collection is finished, and gets the hang of 6. to carry out " exhausting processing ", increased the dead time of application program so again.
To how finding such GC trigger point, a lot of different implementations are arranged in fact equally, provided a GC_schedule interface in the framework, in whole concurrent garbage collector framework, all unify to use this interface to carry out the management of refuse collection thread.Make things convenient for the programmer to realize the dispatching algorithm of different refuse collections by this interface like this.
The second, scheduler needs the situation of disposal system resource exhaustion.At present the Unified frame that realizes exhausts the moment section of generation according to system resource, and can be divided into three kinds to the situation of exhausting: it is in service that one occurs in application program, and the refuse collection thread is not activated; Occur in the tracking phase after the refuse collection thread starts; And the collection clean-up phase after occurring in the refuse collection thread and starting.Scheduler considers from the maximized angle of the operational efficiency of application program, and utilizes the characteristics of concurrent refuse collection, and these three kinds of situations are handled respectively.
(4) for first kind of situation, because the refuse collection thread also is not activated, there is not utilizable information, so can only all threads all be stopped, do once the garbage collection process of " stopping the world ".Allow program continue operation afterwards again.
(5) exhaust tracking phase after appearing at the refuse collection thread and starting when system resource, can be multiplexing during the trace information in so concurrent refuse collection stage.The way of scheduler is to cut off the thread of other application programs, allows the refuse collection thread that tracking phase is finished, and obtains the object trace information earlier.The garbage collection process of doing once " stopping the world " with this object trace information comes releasing memory afterwards.So just having saved a secondary tracking handles.
Exhaust the collection clean-up phase that appears at after the refuse collection thread starts for system resource, handle just more simple.Because tracking before and scale removal process all still are what be suitable for,, allowed the refuse collection thread finish just passable collect clean-up phase as long as the thread of other application programs is cut off in scheduling.
Claims (2)
1. concurrent garbage collector framework based on Java Virtual Machine, it is characterized in that this framework is divided into concurrent algorithm frame and two parts of two modules of scheduler, wherein, described concurrent algorithm frame, mainly carry out variance analysis in design by concurrent garbage collection algorithm to current existence, be convenient to reuse and the module that can not reuse in the final partitioning algorithm, and realize respectively at different characteristics; Again whole concurrent algorithm frame is divided into core algorithm realization, data parallel processing and merging, these three modules of concurrent collection in the existing framework; Described scheduler mainly by scheduling trigger, dispatching algorithm, exhaust and handle three modules and form; This scheduler mainly contains the effect of two aspects:
The first, when decision triggers GC;
The second, the situation of disposal system resource exhaustion; Exhaust the moment section of generation according to system resource, the situation of exhausting is divided into three kinds: it is in service that one occurs in application program, and the refuse collection thread is not activated; Occur in the tracking phase after the refuse collection thread starts; And the collection clean-up phase after occurring in the refuse collection thread and starting; Scheduler is handled respectively these three kinds of situations:
For first kind of situation, only all threads are all stopped, do once the garbage collection process of " stopping the world ", allow program continue operation afterwards again;
When system resource exhausts tracking phase after appearing at the refuse collection thread and starting, scheduler is cut off the thread of other application programs, allows the refuse collection thread that tracking phase is finished, and obtains the object trace information earlier; The garbage collection process of doing once " stopping the world " with this object trace information comes releasing memory afterwards;
Exhaust the collection clean-up phase that appears at after the refuse collection thread starts for system resource, scheduler is only cut off the thread of other application programs, allows the refuse collection thread finish collecting clean-up phase.
2. one kind according to claim 1 based on the concurrent garbage collector framework of Java Virtual Machine, it is characterized in that the transformational relation between each state of framework is:
State is 1.: application program is normally moved, and the operation of at every turn applying for memory headroom all can be called the GC scheduler and be detected whether need to trigger concurrent garbage collection process, and 2. the words that need forward state to;
State is 2.: application program threads is paused respectively, collects the root node on its execution stack, puts into the root node set, and opens the blocker of writing of current thread; After the root node set collection of last thread is finished, start the GC thread and begin to carry out tracing process, forward state to 3.;
State is 3.: 4. the operation that tracing process and application program are concurrent after last GC thread is finished tracing process, forwards state to;
State is 4.: tracing process is finished, and can trigger concurrent collection process when application program has new application internal memory operation, forwards state to 5.;
State is 5.: 1. collection process and application program concurrent running forward state to after last GC thread is finished collection process;
State is 6.: garbage collection process is in state 1., 3., 5. the time, exhaust if the system heap internal memory is employed program, will get the hang of 6. to carry out the system heap internal memory and exhaust processing, forwards state subsequently to 1..
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010595749 CN102023891A (en) | 2010-12-20 | 2010-12-20 | Concurrent garbage collector frame based on Java virtual machine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN 201010595749 CN102023891A (en) | 2010-12-20 | 2010-12-20 | Concurrent garbage collector frame based on Java virtual machine |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN102023891A true CN102023891A (en) | 2011-04-20 |
Family
ID=43865212
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN 201010595749 Pending CN102023891A (en) | 2010-12-20 | 2010-12-20 | Concurrent garbage collector frame based on Java virtual machine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102023891A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102243596A (en) * | 2011-08-09 | 2011-11-16 | 复旦大学 | Adaptive dynamic code unloading method |
| CN102411513A (en) * | 2011-08-10 | 2012-04-11 | 复旦大学 | Garbage collection method for mixed mode execution engine |
| CN102722447A (en) * | 2012-06-06 | 2012-10-10 | 北京航空航天大学 | Incremental track record method of object state in memory garbage collector |
| CN106293878A (en) * | 2016-08-05 | 2017-01-04 | 北京航空航天大学 | A kind of method that dynamic calculation object reuse is worth in JVM |
| CN106293939A (en) * | 2016-08-05 | 2017-01-04 | 北京航空航天大学 | A kind of method of dynamic reuse object in internal memory garbage collector |
| WO2017167061A1 (en) * | 2016-03-29 | 2017-10-05 | 阿里巴巴集团控股有限公司 | Parallel marking processing method and device for garbage collection |
| CN108446162A (en) * | 2018-02-13 | 2018-08-24 | 南京途牛科技有限公司 | Monitor the method and system of JVM Full GC events |
| CN108536537A (en) * | 2018-03-29 | 2018-09-14 | 北京白山耘科技有限公司 | A kind of process resource recovery method and device |
| US10241716B2 (en) | 2017-06-30 | 2019-03-26 | Microsoft Technology Licensing, Llc | Global occupancy aggregator for global garbage collection scheduling |
| US10248562B2 (en) | 2017-06-30 | 2019-04-02 | Microsoft Technology Licensing, Llc | Cost-based garbage collection scheduling in a distributed storage environment |
| CN110291508A (en) * | 2017-02-01 | 2019-09-27 | 微软技术许可有限责任公司 | Garbage collector |
| CN113360274A (en) * | 2020-03-04 | 2021-09-07 | 阿里巴巴集团控股有限公司 | Garbage recycling method and device and computing equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005001695A1 (en) * | 2003-06-30 | 2005-01-06 | Matsushita Electric Industrial Co., Ltd. | Garbage collection system |
| CN1648879A (en) * | 2004-01-28 | 2005-08-03 | 三星电子株式会社 | Method and apparatus for adaptive garbage collection |
| CN101075209A (en) * | 2006-04-28 | 2007-11-21 | 国际商业机器公司 | System and method for collecting garbage in isomerized processor systems |
-
2010
- 2010-12-20 CN CN 201010595749 patent/CN102023891A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2005001695A1 (en) * | 2003-06-30 | 2005-01-06 | Matsushita Electric Industrial Co., Ltd. | Garbage collection system |
| CN1648879A (en) * | 2004-01-28 | 2005-08-03 | 三星电子株式会社 | Method and apparatus for adaptive garbage collection |
| CN101075209A (en) * | 2006-04-28 | 2007-11-21 | 国际商业机器公司 | System and method for collecting garbage in isomerized processor systems |
Non-Patent Citations (1)
| Title |
|---|
| 《复旦大学》 20091231 周寻 基于多核环境下的高效并发性垃圾收集器 第5页至第40页 1-2 , * |
Cited By (20)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102243596A (en) * | 2011-08-09 | 2011-11-16 | 复旦大学 | Adaptive dynamic code unloading method |
| CN102411513A (en) * | 2011-08-10 | 2012-04-11 | 复旦大学 | Garbage collection method for mixed mode execution engine |
| CN102411513B (en) * | 2011-08-10 | 2013-10-16 | 复旦大学 | Garbage collection method for mixed mode execution engine |
| CN102722447A (en) * | 2012-06-06 | 2012-10-10 | 北京航空航天大学 | Incremental track record method of object state in memory garbage collector |
| CN102722447B (en) * | 2012-06-06 | 2015-01-14 | 北京航空航天大学 | Incremental track record method of object state in memory garbage collector |
| US11003578B2 (en) | 2016-03-29 | 2021-05-11 | Banma Zhixing Network (Hongkong) Co., Limited | Method and system for parallel mark processing |
| CN107239406B (en) * | 2016-03-29 | 2021-03-09 | 斑马智行网络(香港)有限公司 | Parallel marking processing method and device for garbage recovery |
| WO2017167061A1 (en) * | 2016-03-29 | 2017-10-05 | 阿里巴巴集团控股有限公司 | Parallel marking processing method and device for garbage collection |
| CN107239406A (en) * | 2016-03-29 | 2017-10-10 | 阿里巴巴集团控股有限公司 | Parallel mark processing method and processing device for garbage reclamation |
| CN106293939B (en) * | 2016-08-05 | 2019-10-18 | 北京航空航天大学 | A method of the dynamic reuse object in memory garbage collector |
| CN106293878B (en) * | 2016-08-05 | 2019-08-27 | 北京航空航天大学 | A method of dynamic computing object reuses value in JVM |
| CN106293939A (en) * | 2016-08-05 | 2017-01-04 | 北京航空航天大学 | A kind of method of dynamic reuse object in internal memory garbage collector |
| CN106293878A (en) * | 2016-08-05 | 2017-01-04 | 北京航空航天大学 | A kind of method that dynamic calculation object reuse is worth in JVM |
| CN110291508A (en) * | 2017-02-01 | 2019-09-27 | 微软技术许可有限责任公司 | Garbage collector |
| US10241716B2 (en) | 2017-06-30 | 2019-03-26 | Microsoft Technology Licensing, Llc | Global occupancy aggregator for global garbage collection scheduling |
| US10248562B2 (en) | 2017-06-30 | 2019-04-02 | Microsoft Technology Licensing, Llc | Cost-based garbage collection scheduling in a distributed storage environment |
| CN108446162A (en) * | 2018-02-13 | 2018-08-24 | 南京途牛科技有限公司 | Monitor the method and system of JVM Full GC events |
| CN108446162B (en) * | 2018-02-13 | 2021-11-26 | 南京途牛科技有限公司 | Method and system for monitoring JVM Full GC event |
| CN108536537A (en) * | 2018-03-29 | 2018-09-14 | 北京白山耘科技有限公司 | A kind of process resource recovery method and device |
| CN113360274A (en) * | 2020-03-04 | 2021-09-07 | 阿里巴巴集团控股有限公司 | Garbage recycling method and device and computing equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102023891A (en) | Concurrent garbage collector frame based on Java virtual machine | |
| CN102073698B (en) | Sample data acquisition method and device for enterprise data warehouse system | |
| CN107203552B (en) | Garbage recovery method and device | |
| US7613753B2 (en) | Platform and method for functional programming (FP) processing | |
| US9189393B2 (en) | Computer, control method of computer, and recording medium | |
| CN108279892A (en) | It is a kind of to split the method, apparatus and equipment that large-scale application service is micro services | |
| CN102170474A (en) | Method and system for dynamic scheduling of virtual resources in cloud computing network | |
| CN103197969A (en) | Distributed timed task control device and method | |
| CN102163161A (en) | Process management method and device thereof | |
| CN102521024B (en) | Job scheduling method based on bioinformation cloud platform | |
| CN101916224A (en) | Test progress abnormality detection method and automatic test control device | |
| CN102768637A (en) | Method and device for controlling test execution | |
| CN105912387A (en) | Method and device for dispatching data processing operation | |
| CN101876918B (en) | Synchronous task sensing method of virtual machine client operation system | |
| CN103440296A (en) | Data query method and device | |
| CN101520748A (en) | Method for testing speed-up ratio of Intel multicore CPU | |
| KR20070088099A (en) | Apparatus and method for managing resource using virtual id under multiple java applications environment | |
| CN111124648B (en) | Multitasking method and device | |
| CN102411513B (en) | Garbage collection method for mixed mode execution engine | |
| CN102541750A (en) | Implementation method and device for data snapshot | |
| CN101635940A (en) | Method and device for measuring program segment execution time | |
| Dai et al. | Research and implementation of big data preprocessing system based on Hadoop | |
| US20230244588A1 (en) | Parallel program scalability bottleneck detection method and computing device | |
| CN110502337B (en) | Optimization system for shuffling stage in Hadoop MapReduce | |
| CN101482831A (en) | Method and equipment for concomitant scheduling of coexisting threading in multi-threading computer system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20110420 |