CN104750562B - Lock processing system in distributed system and method - Google Patents
Lock processing system in distributed system and method Download PDFInfo
- Publication number
- CN104750562B CN104750562B CN201310742719.8A CN201310742719A CN104750562B CN 104750562 B CN104750562 B CN 104750562B CN 201310742719 A CN201310742719 A CN 201310742719A CN 104750562 B CN104750562 B CN 104750562B
- Authority
- CN
- China
- Prior art keywords
- lock
- node
- nodes
- status
- locking
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Hardware Redundancy (AREA)
- Multi Processors (AREA)
Abstract
The present invention provides the lock processing system in a kind of distributed system, which includes multiple nodes, which includes:Lock memory module in each node is set, for storing lock information related with node where it, the lock information includes the lock status of node where it, and the lock status includes idle, pre-locked and locking, and the lock information further includes priority data when the lock status is pre-locked;Lock processing module in each node is set, for node where handling it and the related data of lock and handles interaction data related with lock status between its place node and other nodes;Lock communication module in each node is set, for node where it and the communication between other nodes.Corresponding lock processing system is also provided.The system according to the present invention avoids the problem of the lock processing exception caused by Single Point of Faliure in conventional distributed system with method.
Description
Technical field
The present invention relates to technology related with distributed system, more particularly, to the distribution in distributed system
Lock related technology.
Background technology
The lock implementation method of existing distributed system, mostly uses greatly Centroid system, that is, centered on using host
Node is responsible for the lock request of all nodes by the Centroid, locks and establish, lock revocation, lock overtime control etc..This pattern it is good
It is in simple in structure and be easily achieved and control, but Centroid can only be that a host in other words can only under this structure
It is a node, when Single Point of Faliure occurs for the Centroid, the lock processing for being easy to cause whole system is abnormal or paralyses.
In addition, when lock-based transaction is occurred frequently, each node all goes to access the individual node, causes the individual node pressure to increase, it is likely that become
The bottleneck of whole system.
In addition, there is the scheme using distributed lock system, the pattern of non-stop layer node, but existing this point are used
In cloth scheme, lock information needs to synchronize between each node, this is possible to generate because of lock information synchronization delay, node failure
Etc. the inconsistent problem of lock information caused by reasons, while also limiting the scale of cluster.
Invention content
The present invention provides the lock processing system in a kind of distributed system, can solve the above problems.In the distributed system
Lock processing system be arranged in the distributed system including multiple nodes, which includes:It is arranged in each node
Memory module is locked, for storing lock information related with node where it, the lock information includes the lock status of node where it,
The lock status includes idle, pre-locked and locking, and the lock information further includes priority when the lock status is pre-locked
Data;Lock processing module in each node is set, for node where handling it and locks related data and handles its place section
Interaction data related with lock status between point and other nodes;Lock communication module in each node is set, for section where it
Communication between point and other nodes.
Lock processing system in distributed system according to the present invention, it is preferable that the lock processing module is further
It is arranged to:When lock status of node is pre-locked where it, is sent to other nodes comprising priority data and lock mark
Inquiry request is locked, when the response of other nodes allows locking, the lock status of place node is set to locking, is appointed in the other node
When the response of one node does not allow locking, the lock status of place node is set to the free time;And receive the lock from other nodes
After inquiry request, the lock status of node where inquiring it, when the lock status is idle, generation allows the signal of locking and will
The signal is sent to other nodes, when the lock status is pre-locked, the priority in the lock inquiry request
The priority data of data and node where the lock processing module, and the priority data expression in the lock inquiry request
When priority is high, generates the signal for allowing locking and send the signal to the other nodes for sending out lock inquiry request, instead
It, generates the signal for not allowing locking and sends the signal to the other nodes for sending out lock inquiry request.
Lock processing system in distributed system according to the present invention, it is preferable that the lock processing module is in its institute
When the lock status of node is pre-locked, priority data is generated according to R × 1000+N mod 10000, wherein R is random
Number, N is node serial number, and the node serial number of each node is different.
According to another aspect of the present invention, a kind of lock processing method applied in a distributed system, the distribution are also provided
Formula system includes multiple nodes.The lock processing method of the application in a distributed system includes:The lock status of each node, institute are set
State the held state that lock status shows node to lock;When any node will lock lock, following steps are executed:Inquire the lock of the node
The lock is set to pre-locked by state when the lock status is idle, and it includes lock that priority data and lock lock identify to generate
Inquiry request sends the lock inquiry request to other nodes, response of other nodes to lock inquiry request is received, in the sound
Should be in the case of allowing locking, it is locking to set the lock, and any response is not allow the feelings of locking in the response
Under condition, it is the free time to set the lock;And after any node receives the lock inquiry request from other nodes, following step is executed
Suddenly:The querying node of the lock inquiry request its lock status is received, when its lock status state is locking, generates and sends and does not permit
Perhaps the data locked give other nodes, when the state of its lock status is idle, generate and send the data for allowing locking
To other nodes, when its lock status is pre-locked, priority data in the lock inquiry request with it is described its
The priority data of its node, and it is described lock inquiry request in priority data indicate priority it is high when, generate allow
The signal of locking simultaneously sends the signal to the other nodes for sending out lock inquiry request, conversely, generating does not allow locking
Signal simultaneously sends the signal to the other nodes for sending out lock inquiry request.
Lock processing method in distributed system according to the present invention, it is preferable that node lock status be prelock
Periodically, priority data is generated according to R × 1000+N mod 10000, wherein R is random number, and N is node serial number, and each section
The node serial number of point is different.
According to lock processing scheme provided by the present invention, by information storage related with lock status in distributed system
Each node, and the information is not necessarily to synchronize between each node, but all sections that may hold the lock in direction are initiated by lock
Point inquiry lock status, and solve the problems, such as lock conflict by locking priority data, so as to avoid in the prior art because of concentrate tube
The case where single node of reason lock breaks down and causes system lock processing paralysis.
Description of the drawings
Fig. 1 is the schematic configuration diagram of the lock processing system in exemplary distributed system according to the present invention.
The lock related with the lock node of inquiry request is sent out of Fig. 2 application according to the present invention in a distributed system is handled
The flow chart of method.
Fig. 3 is the lock related with the lock node of inquiry request is received of application according to the present invention in a distributed system
The flow chart of processing method.
Specific implementation mode
Referring now to the schematic example of the attached drawing description present invention, identical drawing reference numeral indicates identical element.Hereafter
Each embodiment of description contributes to those skilled in the art to understand thoroughly the present invention, and is intended to example and unrestricted.Unless otherwise
It limits, term used herein(Including science, technology and industry slang)With general with those skilled in the art in the invention
All over the identical meaning of meaning understood.
Fig. 1 is the schematic configuration diagram of the lock processing system in exemplary distributed system according to the present invention.In this example,
Show four nodes of the distributed system, in concrete application, the data of node can be more or two.Shown in point
Lock processing system in cloth system includes the lock memory module being arranged in each node, related with node where it for storing
Lock information, which includes the lock status of node where it, and the lock status includes idle, pre-locked and locks, described
Lock information further includes priority data when the lock status is pre-locked;Lock processing module in each node is set, for locating
It is related with lock status between node and other nodes where managing its place node and locking related data and handle it to interact number
According to;Communication module in each node is set, for node where communication module and the communication between other nodes.
As shown in Figure 1, node 1 includes the lock memory module 10 being disposed thereon, lock processing module 12 and communication module
14;Node 2 includes the lock memory module 20 being disposed thereon, lock processing module 22 and communication module 24;Node 3 includes setting
Set lock memory module 30, lock processing module 32 and communication module 34 on it;Node 4 includes the latch being disposed thereon
Store up module 40, lock processing module 42 and communication module 44.Lock memory module, lock processing module in each node and lock communication
Module is similar, hereafter will mainly illustrate to lock memory module, lock processing module and lock communication module by taking node 1 as an example.
The storage of lock memory module 10 and the 1 relevant lock information of node of node 1, lock information includes 1 current lock status of node
And lock status be it is pre-locked when further include priority data.The related data with lock that processing module 12 handles node 1 are locked, and
Handle interaction data related with lock status between node 1 and other nodes 2,3 and 4;As an example, lock processing module 12 is from logical
Letter module 14 receives the inquiry request for a certain lock from other nodes and then is made a response to the request, and in node 1
When needing to lock a certain lock, the lock is inquired in the state of node 1 and is further processed according to query result, by communicating mould
Block 14 sends the request to inquiring a certain lock status to other nodes 2,3 and 4 and is handled according to the response of other nodes 2,3 and 4
Lock processing module 12 hereafter will be elucidated in greater detail in the state etc. of node 1 in the lock.In addition, lock processing module 12 is also handled
Revocation, lock overtime control etc. are locked in node and lock related affairs.Communication module 1 is locked to be used to appoint between node 1 and other nodes
The communication of the related data of what and lock, specifically for example including:It receives from other nodes request signal related with lock status
And the request signal is transferred to lock processing module 12, the node 1 that is transmitted by lock processing module 12 is received other nodes is sent out
Lock status inquiry signal and send the signal to other nodes, send by the lock processing module 12 of node 1 generate to it
The response of its node lock inquiry signal and receive other nodes be sent to node 1 lock request response signal and by the letter
Number send lock processing module 12 etc. to.
For convenience of description, a certain lock in the distributed system is hereinafter referred to " A locks ".
The current lock status of each node includes three kinds of " free time ", " pre-locked " and " locking ", wherein " free time " table
Show that A locks are not held by present node, is in idle condition;" pre-locked " indicates the locked A locks of this node, is try to inquire
The lock-out state that other nodes lock A, but not yet receive inquiry result;" locking " indicates that A locks are in locking shape in this node
State and other nodes also agree unanimously that A locks are locked.Specifically, the lock status of node 1 is to indicate " free time " A locks not by node
1 locking;The lock status of node 1 is that " pre-locked " expression A locks are locked by node 1, and node 1 is inquiring other nodes pair
The lock-out state of A locks, but not yet receive response of other nodes to the inquiry;When the lock status of node 1 is " locking ", indicate
A locks are locked by node 1 and other nodes all agree to that A locks are locked by node 1.
When node 1 needs to lock A locks, lock processing module 12 inquires A locks in the current of node 1 to lock memory module 10 first
State is no longer handled if A locks have been in locking, if A locks are currently at idle state, the status maintenance that A is locked
It is changed to " pre-locked ", while generating the lock priority data and A lock status inquiry requests for A locks, and the inquiry request is passed
Lock communication module 14 is given, is for example sent to node 2,3 and 4 in the form of broadcast by lock communication module 14, wherein the inquiry is asked
It asks including the lock priority data locked for A and A lock marks, it is A locks that lock mark, which is intended to be inquired to other node specifications,
The lock priority data shows the priority that node 1 locks A.The communication module 24,34 and 44 of node 2,3 and 4 receives respectively should
Inquiry request, communication module 24 communicate the request to the lock processing module 22 of node 2, and communication module 34 communicates the request to
The lock processing module 32 of node 3, and communication module 44 communicates the request to the lock processing module 42 of node 4.At the lock of node 2
After reason module 22 receives the request, i.e., the lock memory module 20 of query node 2 is to know that A locks the state in node 2;Node
After 3 lock processing module 32 receives the request, i.e., the lock memory module 30 of query node 3 is to know that A locks the shape in node 3
State;After lock processing module 42 receives the request, i.e., the lock memory module 40 of query node 4 is to know that A locks the shape in node 4
State.It is " free time " to lock processing module 22 to obtain A lock current states, therefore generates the signal for allowing locking, and passes through communication module
24 send the signal to node 1.It is " pre-locked " to lock processing module 32 to obtain A locks in the current state of node 3, then further
Query node 3 locks the priority data that A locks in memory module 30, priority data to lock node 3 to A with received
Inquiry request interior joint 1 compares the A priority datas locked, if the priority that node 3 locks A is high, locks processing module 32
It generates the signal for not allowing locking and node 1 is sent the signal to by communication module 34, it is on the contrary then generating allows to lock
Signal simultaneously makes state of the A locks in node 3 " free time " into, and the signal of locking will be allowed to be sent to section by communication module 24
Point 1.It is " locking " to lock processing module 42 to obtain A locks in the current state of node 4, then generating does not allow locking signal, and should
Signal is sent to the communication module 14 of node 1 via communication module 44.If what in this example, node 1 received comes from section
The response of 2,3,4 pairs of lock inquiries of point is to allow to lock, allow locking, do not allow to lock respectively, then node 1 cannot finally lock A
Lock, and the A states locked can be arranged to the free time.
Priority data can be manual setting in each node of distributed system, can also generate at random.It is discussed herein
In example, by lock processing module lock status be it is pre-locked when according to the following formula(1)It generates:
Priority data=R × 1000+N mod 10000(1)
Wherein, R is random number, and N is node serial number, and the number of each node is different.It is noted that the present invention
The algorithm of the calculating priority data provided is only exemplary rather than limitation, can also generate the priority data according to other algorithms.
In addition, in lock processing system as described above, the existing storage mould of each node may be used in each memory module of locking
Block also can be still further arranged, and each lock processing module, which can be implemented separately, also may be implemented in the existing processing component of each node(Such as processing
Device)In, the existing communication module of each node then can be used in each communication module.
Lock processing system in distributed system as described above, will information storage related with lock status in distribution
Each node of system, and the information is not necessarily to synchronize between each node, but by locking initiator(Such as the node in example
1)To being possible to hold the querying node lock status of the lock, and solves the problems, such as lock conflict by locking priority data, thus
Avoid the case where causing system lock processing paralysis due to the single node of centralized management lock breaks down in the prior art.
Fig. 2 is at the lock related with the lock node of inquiry request is sent out of application according to the present invention in a distributed system
The flow chart of reason method, wherein the distributed system includes multiple nodes, such as node 1,2,3 and 4.
In step 200, the lock status of each node is set, which shows held state of the node to lock.In the present invention
Example in, lock status includes idle, locking and pre-locked.Any node in the distributed system, such as node 1 need
Lock A locks(As described above, for convenience, any lock is referred to A locks)When, execute step 210.In step 210, inquiry should
Node, for example, node 1 lock status.In step 212, according to query result, by A locks the state of node 1 be set to it is pre-locked or
No longer carry out subsequent processing;Specifically, when the lock status of node 1 is idle, A locks are set in the lock status of node 1 pre-
Locking then no longer carries out subsequent processing when lock status is locking.By A locks when node 1 is set to pre-locked, that is, execute step
214, generate the lock inquiry request for including priority data and A lock lock marks;Priority data can be for the distribution
Each node in system and manual configuration, also can be when being set as pre-locked, according to formula by the lock status(1)It generates.In step
Rapid 216, the lock inquiry request of generation is sent to other nodes 2,3 and 4.In step 218, the node 1 of lock inquiry request is sent out
Receive the response of the lock inquiry request sent for node 1 from node 2,3 and 4.In step 220, lock inquiry request is sent out
Node 1 sets A locks as locking or free time according to the response received, specifically, the case where the response is to allow locking
Under, it is locking to set the lock, and any response is in the case of not allowing locking in the response, and it is the free time to set the lock.
Fig. 3 is the lock related with the lock node of inquiry request is received of application according to the present invention in a distributed system
The flow chart of processing method.In conjunction with the example above with respect to Fig. 2, illustrate method shown in Fig. 3, also that is, being looked into conjunction with lock is received
The node 2,3 and 4 of request is ask to illustrate method shown in Fig. 3.
Any of the node 2,3 and 4 for receiving lock inquiry request executes this method.In step 300, receives lock and look into
The node 2,3 and 4 for asking request inquire A locks state at which respectively, specifically, the inquiry of node 2 A lock node 2 state,
Node 3 inquires A locks and locks the state in node 4 in state and node 4 the inquiry A of node 3.
In step 302 corresponding signal is generated according to query result;Specifically, when lock status is locking, generation is not permitted
Perhaps the signal locked generates the signal for allowing locking when lock status is idle, when lock status is pre-locked, compares reception
To the lock inquiry request from node 1 in priority data and this node in priority data, in the preferential of lock request
When the priority that level data indicates is high, the signal for allowing locking is generated, conversely, generating the signal for not allowing locking.As an example,
The query result of node 2 is that A locks are idle, and the query result of node 3 is that A locks are pre-locked, and the query result of node 4 is A lock lockings.
Node 2 generates the signal for allowing locking in the case where A locks idle;Node 3 is then further directed to the priority of A locks by node 3
Data are compared the A priority datas locked with the lock inquiry request interior joint 1 received, if node 3 A is locked it is preferential
On the contrary grade is high, then generates not the signal for allowing locking, then generate the signal for allowing locking and change state of the A locks in node 3
At " free time ".In the case where the current state of node 4 is " locking ", the signal for not allowing locking is generated.
In step 304, the signal generated in step 302 is sent to the node 1 for sending out lock inquiry request, that is, node 2 will
The signal for allowing locking generated is sent to node 1, and node 3 allows generation to lock or locking signal is not allowed to be sent to section
Point 1, node 4 would not allow for the signal of locking to be sent to node 1.
Method as described above can be realized as software, hardware or the combination of the rwo.In practical application, side as described above
Method can perform in the lock processing system in conjunction with described in Fig. 1.
Method as described above, makes each node in distributed system be involved in lock management, is intended to lock at it
When a certain lock, you can send out inquiry request to other nodes that may hold the lock, held to the lock with understanding other nodes
State, and lock conflict is solved the problems, such as by priority data.Compared with prior art, lock processing method of the present invention is executed
Distributed system, avoid because lock rely on Centroid due to cause Single Point of Faliure cause whole system lock processing paralysis ask
Topic, while the problem of lock handles bottleneck in distributed system is also solved, improve the high availability of distributed system.
Claims (4)
1. the lock processing system in a kind of distributed system, which includes multiple nodes, which is characterized in that described point
Lock processing system in cloth system includes:
Lock memory module in each node is set, and for storing lock information related with node where it, the lock information includes
The lock status of node where it, the lock status include idle, pre-locked and locking, and the lock information is pre- in the lock status
It further include priority number when locking;
Lock processing module in each node is set, for node where handling it and locks related data and handles its place section
Interaction data related with lock status between point and other nodes;
Lock communication module in each node is set, for node where it and the communication between other nodes, wherein at the lock
Reason module is further arranged to:
When lock status of node is pre-locked where it, sends the lock comprising priority data and lock mark to other nodes and look into
Request is ask,
When the response of other nodes allows locking, the lock status of place node is set to locking,
When any node response does not allow locking in the other node, the lock status of place node is set to the free time;And
After receiving the lock inquiry request from other nodes, the lock status of node where inquiring it,
When the lock status is idle, generates the signal for allowing locking and sends the signal to other nodes,
When the lock status is pre-locked, the priority data in the lock inquiry request and lock processing module place
The priority data of node, and it is described lock inquiry request in priority data indicate priority it is high when, generation allow to lock
Fixed signal and the other nodes for sending out lock inquiry request are sent the signal to, conversely, generating the letter for not allowing locking
Number and send the signal to send out lock inquiry request other nodes.
2. the lock processing system in distributed system as described in claim 1, which is characterized in that the lock processing module is at it
When the lock status of place node is pre-locked, priority data is generated according to R × 1000+N mod 10000, wherein R is random
Number, N is node serial number, and the node serial number of each node is different.
3. a kind of lock processing method of application in a distributed system, which includes multiple nodes, which is characterized in that
The application lock processing method in a distributed system includes:
The lock status of each node is set, and the lock status shows held state of the node to lock;
When any node will lock lock, following steps are executed:
The lock status for inquiring the node, when the lock status is idle, which is set to it is pre-locked,
The lock inquiry request for including priority number and lock lock mark is generated,
The lock inquiry request is sent to other nodes,
Response of other nodes to lock inquiry request is received,
In the case where the response is to allow locking, it is locking to set the lock, and any response is not in the response
In the case of allowing locking, it is the free time to set the lock;And
After any node receives the lock inquiry request from other nodes, following steps are executed:
Receive the querying node of the lock inquiry request its lock status;
When its lock status is locking, the data for not allowing to lock are generated and sent to other nodes;
When its lock status is idle, the data for allowing to lock are generated and sent to other nodes;
When the lock status is pre-locked, the priority number locked in inquiry request is preferential with other nodes
Series, and it is described lock inquiry request in priority number indicate priority it is high when, generate allow locking signal and should
Signal is sent to the other nodes for sending out lock inquiry request, conversely, generating the signal for not allowing locking and sending out the signal
Give the other nodes for sending out lock inquiry request.
4. the lock processing method of application as claimed in claim 3 in a distributed system, which is characterized in that the priority exists
It is generated by R × 1000+N mod10000 when the lock status is pre-locked, wherein R is random number, and N is node serial number, and each section
The node serial number of point is different.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310742719.8A CN104750562B (en) | 2013-12-30 | 2013-12-30 | Lock processing system in distributed system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310742719.8A CN104750562B (en) | 2013-12-30 | 2013-12-30 | Lock processing system in distributed system and method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN104750562A CN104750562A (en) | 2015-07-01 |
CN104750562B true CN104750562B (en) | 2018-09-21 |
Family
ID=53590303
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310742719.8A Active CN104750562B (en) | 2013-12-30 | 2013-12-30 | Lock processing system in distributed system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104750562B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105069008A (en) * | 2015-07-03 | 2015-11-18 | 曙光信息产业股份有限公司 | Distributed system data processing method and apparatus |
CN106293954A (en) * | 2016-08-08 | 2017-01-04 | 浪潮(北京)电子信息产业有限公司 | A kind of High Availabitity service management based on distributed lock |
CN106484542B (en) * | 2016-09-06 | 2020-05-19 | 华为技术有限公司 | Method and device for processing overlapping node event in distributed system |
CN106775974B (en) * | 2016-12-07 | 2019-12-10 | 国云科技股份有限公司 | Method for realizing distributed priority queuing lock |
CN107517277B (en) * | 2017-10-09 | 2020-05-19 | 苏州浪潮智能科技有限公司 | Method and device for realizing sanlock |
CN109697127B (en) * | 2017-10-24 | 2020-12-01 | 华为技术有限公司 | Method and device for locking access operation of shared resource |
CN112035269B (en) * | 2020-09-10 | 2021-06-01 | 湖南文盾智链科技有限公司 | Multi-node locking method for distributed system |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101256510A (en) * | 2008-04-11 | 2008-09-03 | 中兴通讯股份有限公司 | Cluster system and method for implementing centralized management thereof |
CN102412988A (en) * | 2011-11-14 | 2012-04-11 | 浪潮(北京)电子信息产业有限公司 | Service information system and method for realizing continuous operation by using same |
-
2013
- 2013-12-30 CN CN201310742719.8A patent/CN104750562B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101256510A (en) * | 2008-04-11 | 2008-09-03 | 中兴通讯股份有限公司 | Cluster system and method for implementing centralized management thereof |
CN102412988A (en) * | 2011-11-14 | 2012-04-11 | 浪潮(北京)电子信息产业有限公司 | Service information system and method for realizing continuous operation by using same |
Also Published As
Publication number | Publication date |
---|---|
CN104750562A (en) | 2015-07-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104750562B (en) | Lock processing system in distributed system and method | |
WO2015096606A1 (en) | Network device, cluster storage system and distributed lock management method | |
US8671151B2 (en) | Maintaining item-to-node mapping information in a distributed system | |
US9563673B2 (en) | Query method for a distributed database system and query apparatus | |
WO2015062444A1 (en) | System and method for creating a distributed transaction manager supporting repeatable read isolation level in a mpp database | |
EP3208998B1 (en) | Method, apparatus and system for executing distributed transaction resources | |
US20190075084A1 (en) | Distributed Lock Management Method, Apparatus, and System | |
KR20160051994A (en) | Method and Apparatus for Supplying Key to Multiple Devices in Quantum Key Distribution System | |
US20080209007A1 (en) | Methods, systems, and computer program products for accessing data associated with a plurality of similarly structured distributed databases | |
WO2017152860A1 (en) | Heartbeat information sending method and device, and heartbeat sending node | |
US7974218B2 (en) | Intersection-based configuration management | |
US20210075711A1 (en) | Cluster wide packet tracing | |
JP2010510578A (en) | SAS zone group permission table version identifier | |
CN108984639B (en) | Data processing method and device for server cluster | |
KR20110110223A (en) | Selective database replication | |
US8478803B2 (en) | Management of logical statements in a distributed database environment | |
US9020888B1 (en) | Data replicating systems and data replication methods | |
WO2017012383A1 (en) | Service registration method, usage method and relevant apparatus | |
CN103440345A (en) | Distributed database extension method and distributed database extension system based on relational database | |
EP3128715B1 (en) | Resource creation method and apparatus | |
JP5110998B2 (en) | Distribution device, communication system, and communication method | |
CN108139927A (en) | The routing based on action of affairs in online transaction processing system | |
CN106909197B (en) | Virtualization host time management method and virtualization host system | |
US11526560B2 (en) | Backwards-compatible method for retrieving data stored in a database using versioned reference perspectives of edges | |
WO2016124075A1 (en) | Method and device for controlling management mode of optical network unit in passive optical network |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |