CN106817231A - A kind of Virtual Cluster rollback method and system based on multicasting technology - Google Patents
A kind of Virtual Cluster rollback method and system based on multicasting technology Download PDFInfo
- Publication number
- CN106817231A CN106817231A CN201510890303.XA CN201510890303A CN106817231A CN 106817231 A CN106817231 A CN 106817231A CN 201510890303 A CN201510890303 A CN 201510890303A CN 106817231 A CN106817231 A CN 106817231A
- Authority
- CN
- China
- Prior art keywords
- packet
- virtual machine
- page
- multicast
- data
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F12/00—Accessing, addressing or allocating within memory systems or architectures
- G06F12/02—Addressing or allocation; Relocation
- G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/44—Arrangements for executing specific programs
- G06F9/455—Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
- G06F9/45504—Abstract machines for programme code execution, e.g. Java virtual machine [JVM], interpreters, emulators
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L12/00—Data switching networks
- H04L12/02—Details
- H04L12/16—Arrangements for providing special services to substations
- H04L12/18—Arrangements for providing special services to substations for broadcast or conference, e.g. multicast
- H04L12/185—Arrangements for providing special services to substations for broadcast or conference, e.g. multicast with management of multicast group membership
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/16—Multipoint routing
Landscapes
- Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Theoretical Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a kind of Virtual Cluster rollback method based on multicasting technology and system.This method is:1) storage cluster identifies the data identical virtual machine page from snapshot document, and then the corresponding relation according to virtual machine and physical machine determines the virtual machine page for needing multicast or unicast, and packet encapsulation is carried out to the virtual machine page;2) storage cluster judges whether current packet to be sent is multicast packet, if multicast packet, then notify the multicast packet to be sent to physical machine be added in a multicast group, then the multicast packet is sent in the multicast group;Physical machine is parsed to the multicast packet for receiving, and the virtual machine page in multicast packet is returned into corresponding address;Correspondence physical machine is sent to unicast fashion if unicast packet, the virtual machine page in the packet is returned to corresponding address by physical machine.The present invention can carry out quick rollback to Virtual Cluster, and reduce transmitted data amount.
Description
Technical field
The invention belongs to technical field of virtualization, it is related to virtual machine snapshot rollback technology, and in particular to be based on multicast skill to one kind
The Virtual Cluster rollback method and system of art.
Background technology
At present, calculated on processing platform on a large scale as data center or cloud computing, in order to meet Distributed Application journey
The demand of ordered pair disposal ability, virtual machine is with cluster form disposing application program.However, frequently failure is past in extensive platform
Toward cause cluster in application crash.In order to ensure application availability, snapshot rollback technology is suggested, its in program just
Be saved into the run time behaviour of Virtual Cluster in disk by the normal execution stage, and after the failure of cluster program, these data will be read
To recover the execution of cluster.
Fig. 1 shows the typical architecture of data center.Storage device and server are often placed respectively, and wherein storage device passes through
Storage network connection, and server is then managed by three layer data central site networks.Under this framework, Virtual Cluster snapshot
File is stored in storage device, and in rollback, storage file is final to send out by storage network transmission to three layer data central site networks
Give server.Virtual machine on server then reads data from snapshot document, is loaded into its internal memory and equipment, then extensive
It is multiple to perform.Because virtual machine run time behaviour data volume is big, therefore virtual machine snapshot file is accordingly larger.On the other hand, virtually
Cluster is because comprising multiple virtual machines, its snapshot document will be the several times of unit.This produces net higher in Virtual Cluster rollback
Network expense, while in the limited network bandwidth, introducing rollback higher and postponing.Therefore, how in the limited network bandwidth,
Network overhead and the delay of cluster rollback are reduced, is one of key issue in Virtual Cluster rollback technology.
The invention discloses a kind of Virtual Cluster rollback method based on multicast and system, it utilizes multicasting technology that identical data is same
When be transferred on multiple servers, so as to effectively reduce transmitted data on network amount, and accelerate rollback time.
The content of the invention
It is a primary object of the present invention to propose a kind of Virtual Cluster rollback method based on multicasting technology and system, can be to virtual
Cluster carries out quick rollback, and reduces the transmitted data amount produced in rollback process.
The technical scheme is that:
A kind of Virtual Cluster rollback method based on multicasting technology, its step is:
1) storage cluster identifies the data identical virtual machine page from snapshot document, then right with physical machine according to virtual machine
Should be related to and determine the virtual machine page for needing multicast or unicast, packet encapsulation is carried out to the virtual machine page;Wherein,
The virtual machine page that data are identical and corresponding virtual machine recovers in same physical machine is defined as into mode of unicast to pass
It is defeated, the virtual machine page that data content is identical and corresponding virtual machine recovers in different physical machines is defined as multicast side
Formula is transmitted;
2) storage cluster judges whether current packet to be sent is multicast packet, and if multicast packet, then notifying should
Multicast packet to be sent to physical machine be added in a multicast group, the multicast packet is then sent to the multicast group
In;Physical machine is parsed to the multicast packet for receiving, and the virtual machine page in multicast packet is returned to accordingly
Address;Correspondence physical machine is sent to unicast fashion if unicast packet, physical machine is by the void in the packet
The plan machine page returns to corresponding address;Wherein, the physical machine in multicast group be correspondence multicast packet to be sent to thing
Reason machine set.
Further, it is to the form that the data identical virtual machine page is carried out after packet encapsulation:[number of pages, { virtual machine volume
Number i, page number j } ..., { virtual machine numbering k, page number h }, content of pages].
Further, the step 2) in, storage cluster preferentially sends multicast packet, then sends unicast packet.
Further, for unicast packet, the preferential transmission content of storage cluster is identical and is sent to the same unicast of physical machine
Packet.
Further, the storage cluster adds multicast group and exits multicast group by control command notice physical machine, safeguards multicast
Physical machine in group be correspondence multicast packet to be sent to physical machine set.
A kind of Virtual Cluster rollback system based on multicasting technology, it is characterised in that including storage cluster and server cluster, its
A rollback client, control client and data client are disposed on each storage device end in middle storage cluster;Server set
Rollback service end, control service end and data service end are disposed in each physical machine in group;
The rollback client, be responsible for the data identical virtual machine page is identified from snapshot document, then according to virtual machine with
The corresponding relation of physical machine determines the virtual machine page for needing multicast or unicast;
The control client, is responsible for sending control command, notifies that the physical machine in server cluster adds multicast group and exits group
Broadcast group;
The data client, is responsible for carrying out packet encapsulation and transmission to the virtual machine page, wherein, data are identical and corresponding
The virtual machine page that recovers in same physical machine of virtual machine be encapsulated as unicast packet, data content is identical and corresponding
The virtual machine page that virtual machine recovers in different physical machines is encapsulated as multicast packet;
The rollback service end, for by controlling service end and data service end to receive control command and packet, Yi Jigen
Multicast group is added or exits according to the control command control physical machine for receiving;Packet to receiving is parsed, by packet
The virtual machine page returns to corresponding address.
Compared with prior art, the positive effect of the present invention is:
The present invention can carry out quick rollback to Virtual Cluster, and reduce the transmitted data amount produced in rollback process.The present invention
Good effect embody in the following areas:1) by merging page data, so as to the data volume of transmission will be needed to greatly reduce.It is special
It is not under conditions of there is more identical data in Virtual Cluster between virtual machine, will have substantial amounts of data identical and merge.2) it is right
In needing to be transferred to many pages of physical machine, transmitted by multicasting technology, only sending portion just can be by Simultaneous Transmission of Data
Onto many physical machines, it is to avoid the repetition transmission of same page.
Brief description of the drawings
Fig. 1 is data center architecture figure;
Fig. 2 is the page transmission schematic diagram based on multicast;
Fig. 3 is Virtual Cluster rollback system Organization Chart;
Fig. 4 is Virtual Cluster rollback flow chart.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, the present invention is done with specific embodiment below in conjunction with the accompanying drawings
Further illustrate.
We are mainly solved when Virtual Cluster needs rollback, because the network overhead for reading multiple snapshot documents introducings is big and rollback is prolonged
Problem big late.Our solution is as follows:
Virtual machine in Virtual Cluster often configures identical operating system, development library and application program, and this causes virtual machine in fortune
During row distributed program, often possesses identical data content.Therefore, by recognize virtual machine internal and between same page,
Same page is merged and is transmitted, the data volume of transmission required for can effectively reducing.Meanwhile, during Virtual Cluster rollback,
Virtual machine is often placed in different physical machines.Therefore, for the transmission of same page, we can by multicasting technology,
Come only transmit a data and while transmitting data on many physical nodes.As shown in Fig. 2 snapshot 1, snapshot 2 and fast
It is virtual machine 1, virtual machine 2 and the corresponding snapshot document of virtual machine 3 according to 3.In rollback, all deposited in snapshot 1 and snapshot 2
In page A, meanwhile, virtual machine 1 and virtual machine 2 start on main frame 1, therefore A will be merged and be transmitted with mode of unicast.
For page B, it is present in three snapshot documents, because virtual machine 3 starts on main frame 2, and it is virtual machine 1, virtual
Machine 2 starts on main frame 1, therefore page B needs to be sent to two physical machines, will be transmitted by multicasting technology.For the page
D and page E, it only has portion, therefore will be transmitted in the way of unicast.
Rollback flow.The rollback of Virtual Cluster is main to be implemented at storage device end, and storage device end is processed snapshot document,
Identify which page should unicast or multicast, the page encapsulated and is sent to thing by the page for needing multicast by udp protocol
Reason machine, and then sent using TCP for the page of unicast.The forwarding of packet is completed by data center network after encapsulation.Tool
For body, the rollback of Virtual Cluster is divided into three phases:Preparatory stage, multicast stage and unicast stage.
1) preparatory stage is mainly responsible for reading snapshot document, and identifies the data identical page by page contrast.Meanwhile,
Rollback position based on virtual machine specified by user, i.e. virtual machine recover to perform in which platform physical machine, judge to need multicast or
The page of unicast.Specifically, if two page (adhering to two virtual machines separately) data are identical, and two virtual machines are at same
Physical machine is recovered to perform, then the two pages will be transmitted in the way of unicast, the page A in such as Fig. 2.If two virtual
Machine recovers to perform respectively in two physical machines, then two pages are transmitted in the way of multicast, the page C in such as Fig. 2.For
The data identical page, no matter it is unicast or multicast, and be encapsulated into the page in packet by we.Data packet format is the [page
Number, { virtual machine numbering 1, page number 1 }, { virtual machine numbering 2, page number 2 }, { virtual machine numbering 2, page number 6 } ...,
{ virtual machine numbering k, page number k }, content of pages].
2) the multicast stage, the page of multicast or unicast is needed due to have identified, this stage is mainly responsible for needing multicast page
The transmission in face.For the page of multicast, step 1) in same page is encapsulated into packet.Simultaneously as virtual machine
(typically specified by user) with known to the corresponding relation of physical machine, therefore corresponding relation of the virtual machine page with physical machine can be known,
And then the physical machine that the data bag should be sent to can be obtained.The stage is responsible for notifying that physical machine is added in multicast group, finally by number
It is sent in multicast group according to bag.During next data bag pending, recalculate to be sent to physical machine, notify these physical machines
It is added in multicast group, packet is then issued into newly-generated multicast group.The process will be repeated, until institute in need group
The page broadcast all sends.Recipient, i.e., after physical machine receives multicast packets, packet is parsed, and foundation will be in the thing
The virtual machine (such as virtual machine numbering 1, virtual machine numbering 2) recovered on reason machine, obtains address that the page should fill (for example { virtually
Machine numbering 1, page number 2 }, { virtual machine numbering 2, page number 4 }, { virtual machine numbering 2, page number 7 }, virtual machine numbering 2,
Page number 9 } ... { virtual machine numbering 2, page number 14632 }), finally copy content of pages to corresponding address.
3) the unicast stage.The process will find the page not sent in snapshot document, and these pages include two classes:One class is content
It is identical, but the same page of physical machine is sent to, the page A in such as Fig. 2;Another kind of is the unique page of content, i.e.,
It is same without other page data contents, the page E in such as Fig. 2.This two classes page is all sent to thing by unicast technique
Reason machine.The first kind page is processed in transmission process first, the data packet format of the first kind page is identical with multicast packet form,
The treatment of recipient is also identical, and difference is sender and need not build multicast group (because only that a physical machine).Equations of The Second Kind
The form of packet is then simple, and it is [1 (the only one of which page), virtual machine numbering, page number, content of pages], and recipient connects
To after the packet, in copying content of pages the page number of the virtual machine specified in packet to.
After the completion of 3 stages, all of page has all been sent completely.The page is all being copied to recipient correspondence virtual machine
Corresponding page number after, recover virtual machine perform.
System architecture.Virtual Cluster rollback system framework based on multicast disclosed by the invention is as shown in Figure 3.Rollback client end
At storage device end, it is responsible for reading data in snapshot document, and identifies same page using scheduler, then according to use for administration
Virtual machine and physical machine corresponding relation that family is specified, which platform physical machine this is sent to calculate the packet after page-out encapsulation.Control visitor
Control command is responsible for sending in family end, and data client PkDataClien is responsible for encapsulating and sends packet, including multicast packet with
Unicast packet.Rollback service end is deployed in each physics generator terminal, and it is by controlling service end and data service end to receive control
Order and packet.Control command is mainly responsible for the generation of multicast group, that is, notify that physical machine adds multicast group and exits multicast group.
Packet includes:Unicast packet, multicast packet.
Cluster rollback agreement.As shown in figure 4, wherein there are two roles in the rollback agreement of Virtual Cluster:It is deployed in storage device
Rollback client, and be deployed in the rollback service end RBServer on every server, idiographic flow is as follows.
1st, keeper initiate rollback order " cluster rollback ";
2nd, rollback client to all rollback transmitting terminals send current data packet it is corresponding " multicast group information ".Rollback is serviced
After end receives control bag, if it needs to receive packet, it is added in group;If it in multicast group and is not needed
Packet is received, then exits multicast group, if ignoring the control command not in multicast group and if packet need not being received;
3rd, rollback service end is after adding or exiting multicast group, send to rollback client " multicast group acknowledge ", inform back
Rolling client executing is completed;
4th, rollback client delivers a packet to multicast group in multicast form;
5th, 2-4 steps are repeated, until the page for needing multicast is sent;
6th, " multicast terminates " to notify that multicast is completed rollback service end packet the transmission of rollback client;
7th, the page that rollback client will not send in snapshot document is packaged, and packet is sent to mode of unicast
Corresponding rollback service end;
8th, after being sent completely, " unicast terminates " to notify that the unicast of rollback service end packet is completed the transmission of rollback client;
9th, rollback service end safeguards an array to record the page number for receiving, and according to the memory size of virtual machine,
The unreceived page is searched, the page that then these are lacked to rollback client request.Such as virutal machine memory size is 40K (10
The individual page), if service end has only received certain page of virtual machine 1,2,5,6,8,9,10, then the page 3,4,7 is not yet received,
Missing is denoted as, it is necessary to the packet lost on last stage to rollback client request.
10th, rollback client is searched corresponding after the request missing page for receiving certain rollback service end from snapshot document
The page, is packaged into packet, and the rollback service end is then sent in the form of unicast;
11st, rollback service end replys " data validation " information after all of page is received;
12nd, rollback client receive that all rollback service ends reply " data validation " and after, inform that keeper virtually collects
Group's rollback is completed.
Claims (9)
1. a kind of Virtual Cluster rollback method based on multicasting technology, its step is:
1) storage cluster identifies the data identical virtual machine page from snapshot document, then right with physical machine according to virtual machine
Should be related to and determine the virtual machine page for needing multicast or unicast, packet encapsulation is carried out to the virtual machine page;Wherein,
The virtual machine page that data are identical and corresponding virtual machine recovers in same physical machine is defined as into mode of unicast to pass
It is defeated, the virtual machine page that data content is identical and corresponding virtual machine recovers in different physical machines is defined as multicast side
Formula is transmitted;
2) storage cluster judges whether current packet to be sent is multicast packet, and if multicast packet, then notifying should
Multicast packet to be sent to physical machine be added in a multicast group, the multicast packet is then sent to the multicast group
In;Physical machine is parsed to the multicast packet for receiving, and the virtual machine page in multicast packet is returned to accordingly
Address;Correspondence physical machine is sent to unicast fashion if unicast packet, physical machine is by the void in the packet
The plan machine page returns to corresponding address;Wherein, the physical machine in multicast group be correspondence multicast packet to be sent to thing
Reason machine set.
2. the method for claim 1, it is characterised in that the lattice after packet encapsulation are carried out to the data identical virtual machine page
Formula is:[number of pages, { virtual machine numbering i, page number j } ..., { virtual machine numbering k, page number h }, content of pages].
3. the method for claim 1, it is characterised in that the step 2) in, storage cluster preferentially sends multicast packet,
Then unicast packet is sent.
4. the method as described in claim 1 or 2 or 3, it is characterised in that for unicast packet, in storage cluster preferentially sends
Appearance is identical and is sent to the same unicast packet of physical machine.
5. the method for claim 1, it is characterised in that the storage cluster notifies that physical machine adds multicast by control command
Group and exit multicast group, safeguard the physical machine in multicast group be correspondence multicast packet to be sent to physical machine set.
6. a kind of Virtual Cluster rollback system based on multicasting technology, it is characterised in that including storage cluster and server cluster, wherein
A rollback client, control client and data client are disposed on each storage device end in storage cluster;Server set
Rollback service end, control service end and data service end are disposed in each physical machine in group;
The rollback client, is responsible for identifying the data identical virtual machine page from snapshot document, then according to virtual machine
Being determined with the corresponding relation of physical machine needs the virtual machine page of multicast or unicast;
The control client, is responsible for sending control command, notifies that the physical machine in server cluster adds multicast group and exits
Multicast group;
The data client, is responsible for carrying out packet encapsulation and transmission to the virtual machine page, wherein, data are identical and right
The virtual machine page that the virtual machine answered recovers in same physical machine is encapsulated as unicast packet, by data content it is identical and
The virtual machine page that corresponding virtual machine recovers in different physical machines is encapsulated as multicast packet;
The rollback service end, for receiving control command and packet by controlling service end and data service end, and
Control command control physical machine according to receiving adds or exits multicast group;Packet to receiving is parsed, by data
The virtual machine page in bag returns to corresponding address.
7. system as claimed in claim 6, it is characterised in that the lattice after packet encapsulation are carried out to the data identical virtual machine page
Formula is:[number of pages, { virtual machine numbering i, page number j } ..., { virtual machine numbering k, page number h }, content of pages].
8. system as claimed in claims 6 or 7, it is characterised in that the data client preferentially sends multicast packet, then
Send unicast packet.
9. system as claimed in claims 6 or 7, it is characterised in that for unicast packet, the data client preferentially sends
Content is identical and is sent to the same unicast packet of physical machine.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510890303.XA CN106817231A (en) | 2015-12-07 | 2015-12-07 | A kind of Virtual Cluster rollback method and system based on multicasting technology |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510890303.XA CN106817231A (en) | 2015-12-07 | 2015-12-07 | A kind of Virtual Cluster rollback method and system based on multicasting technology |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106817231A true CN106817231A (en) | 2017-06-09 |
Family
ID=59106275
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510890303.XA Pending CN106817231A (en) | 2015-12-07 | 2015-12-07 | A kind of Virtual Cluster rollback method and system based on multicasting technology |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106817231A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112073209A (en) * | 2019-06-10 | 2020-12-11 | 中兴通讯股份有限公司 | Data packet processing method and device |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103593259A (en) * | 2013-10-16 | 2014-02-19 | 北京航空航天大学 | Snapshot method and snapshot device for memories virtual clusters |
CN103595768A (en) * | 2013-10-28 | 2014-02-19 | 国家电网公司 | Method for realizing configuration synchronization between virtualized devices |
-
2015
- 2015-12-07 CN CN201510890303.XA patent/CN106817231A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103593259A (en) * | 2013-10-16 | 2014-02-19 | 北京航空航天大学 | Snapshot method and snapshot device for memories virtual clusters |
CN103595768A (en) * | 2013-10-28 | 2014-02-19 | 国家电网公司 | Method for realizing configuration synchronization between virtualized devices |
Non-Patent Citations (1)
Title |
---|
BIN SHI 等: "Mercurial: A Traffic-saving Roll Back System for Virtual Machine Cluster", 《2014 IEEE/ACM 7TH INTERNATIONAL CONFERENCE ON UTILITY AND CLOUD COMPUTING》 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112073209A (en) * | 2019-06-10 | 2020-12-11 | 中兴通讯股份有限公司 | Data packet processing method and device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111539726B (en) | Block chain consensus system and method | |
CN108040034A (en) | Distributing real communication middleware | |
CN104092719B (en) | Document transmission method, device and distributed cluster file system | |
CN113347164B (en) | Block chain-based distributed consensus system, method, device and storage medium | |
CN111327414A (en) | Block chain consensus method and system, computer storage medium and electronic device | |
CN103780615B (en) | Sharing method of client conversation data among multiple servers | |
CN105939297A (en) | TCP message reassembling method and TCP message reassembling device | |
CN109254854A (en) | Asynchronous invoking method, computer installation and storage medium | |
EP3235210A1 (en) | Methods and first network node for managing a stream control transmission protocol association | |
CN105847386A (en) | Live broadcasting stream processing method, device and system | |
CN108306852A (en) | A kind of message-oriented middleware system and method based on simple binary coding | |
CN105190530A (en) | Transmitting hardware-rendered graphical data | |
CN109062681A (en) | A kind of execution method, system, device and the storage medium of periodic cycle task | |
CN106059936A (en) | Method and device for multicasting files in cloud system | |
CN108023968A (en) | A kind of session information synchronous method, device and equipment | |
CN103560939A (en) | Scheme for avoiding deadlock in multi-ring interconnect, with additional application to congestion control | |
Collins et al. | Online payments by merely broadcasting messages (extended version) | |
CN106817231A (en) | A kind of Virtual Cluster rollback method and system based on multicasting technology | |
Sussman et al. | Optimistic virtual synchrony | |
CN117061072B (en) | Message transmission method, device, equipment and storage medium | |
AT512665A1 (en) | Method and apparatus for forming software fault containment units in a distributed real-time system | |
CN102123079B (en) | Method and device for processing transaction message | |
CN106685854A (en) | Data transmission method and system | |
CN110445580A (en) | Data transmission method for uplink and device, storage medium, electronic device | |
CN111538588A (en) | Distributed data processing method and related device |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170609 |