CN108667641A - Distributed system upgrade method, distributed system and updating and management equipment - Google Patents
Distributed system upgrade method, distributed system and updating and management equipment Download PDFInfo
- Publication number
- CN108667641A CN108667641A CN201710200566.2A CN201710200566A CN108667641A CN 108667641 A CN108667641 A CN 108667641A CN 201710200566 A CN201710200566 A CN 201710200566A CN 108667641 A CN108667641 A CN 108667641A
- Authority
- CN
- China
- Prior art keywords
- node
- upgraded
- upgrade
- calculate
- calculate node
- 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
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0803—Configuration setting
- H04L41/0813—Configuration setting characterised by the conditions triggering a change of settings
- H04L41/082—Configuration setting characterised by the conditions triggering a change of settings the condition being updates or upgrades of network functionality
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/22—Arrangements for detecting or preventing errors in the information received using redundant apparatus to increase reliability
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/08—Configuration management of networks or network elements
- H04L41/0893—Assignment of logical groups to network elements
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Stored Programmes (AREA)
Abstract
An embodiment of the present invention provides a kind of distributed system upgrade method, distributed system and updating and management equipment, first aspect, in this programme, calculate node is upgraded in batches, namely part calculate node is upgraded every time, other calculate nodes can continue to offer service, in this way, system can continue to offer service while upgrading.Second aspect, calculate node to be upgraded is before upgrading, the service of self-operating or task can be switched to other promoted completions or this time without in the calculate node of upgrading, in this way, can distributed system not influenced by upgrading, external offer service is provided, the service quality that distributed system externally provides is improved.
Description
Technical field
The present invention relates to field of computer technology, more particularly to a kind of distributed system upgrade method, distributed system
And updating and management equipment.
Background technology
Distributed system generally include more communication connection node devices, these node devices can be divided into host node,
Slave node and calculate node, host node and slave node can be scheduled calculate node management, and calculate node can provide
Various services.
When upgrading to distributed system, it usually needs upgrade successively to each node device in system.Mesh
Before, distributed system interior joint equipment is more and more, and function also becomes increasingly complex so that the updating operation of distributed system expends
Time it is increasingly longer.
In general, distributed system cannot provide service during upgrading, that is to say, that the upgrading of distributed system is grasped
Make the long normal use that can influence system.
Invention content
The embodiment of the present invention is designed to provide a kind of distributed system upgrade method, distributed system and updating and management
Equipment can continue to offer service while upgrading.
In order to achieve the above objectives, the embodiment of the invention discloses a kind of distributed system upgrade method, the distributed systems
System includes：Host node, slave node and calculate node；The method includes：
Upgrade command is sent to slave node, so that slave node carries out updating operation；
After the completion of detecting slave node upgrading, calculate node to be upgraded is determined；Wherein, identified calculating section to be upgraded
Point quantity is less than the total quantity of calculate node；
Upgrade command is sent to identified calculate node to be upgraded, so that calculate node to be upgraded carries out updating operation；
Judge whether the quantity for the calculate node that upgrading is completed reaches predetermined threshold value, the predetermined threshold value is saved no more than calculating
The total quantity of point；
If not, returning to the step of executing determination calculate node to be upgraded；
If so, the host node is switched to slave node；New host node is determined in the slave node that upgrading is completed；It will
The calculate node that upgrading is completed is connected to the new host node；The slave node that does not upgrade into the system and/or do not upgrade
Calculate node send upgrade command so that the calculate node not upgraded and/or slave node carry out updating operation.
Optionally, before the transmission upgrade command to slave node, can also include：
By switching scheduling strategy, slave node and the connection of host node and the connection of slave node and calculate node are disconnected；
The method can also include：
After the completion of whole node upgradings in detecting the system, by switching scheduling strategy, slave node and master are established
The connection of the connection of node and slave node and calculate node.
Optionally, after disconnecting the connection of slave node and calculate node, can also include：
The calculate node that will be switched off connection is connected to host node.
Optionally, the method can also include：
After the completion of whole node upgradings in detecting the system, according to default concatenate rule, calculate node is connected
To corresponding host node or slave node.
Optionally, described to send upgrade command to slave node, so that slave node carries out updating operation, including：
The first upgrade procedure is obtained, first upgrade procedure is sent to slave node, so that slave node utilizes described the
One upgrade procedure is upgraded；
Alternatively, upgrade command is sent to slave node, so that slave node obtains the first upgrade procedure according to the upgrade command,
And upgraded using first upgrade procedure；
It is described to send upgrade command to identified calculate node to be upgraded, so that calculate node to be upgraded carries out upgrading behaviour
Make, including：
The second upgrade procedure is obtained, second upgrade procedure is sent to calculate node to be upgraded, so that meter to be upgraded
Operator node is upgraded using second upgrade procedure；
Alternatively, upgrade command is sent to calculate node to be upgraded, so that calculate node to be upgraded is according to the upgrade command
The second upgrade procedure is obtained, and is upgraded using second upgrade procedure；
The slave node not upgraded into the system and/or the calculate node not upgraded send upgrade command, so that
The calculate node not upgraded and/or slave node carry out updating operation, including：
Third upgrade procedure is obtained, the third upgrade procedure is sent to the slave node not upgraded and/or is not upgraded
Calculate node, so that the calculate node not upgraded and/or slave node are upgraded using the third upgrade procedure；
Alternatively, upgrade command is sent to calculate node to be upgraded, so that the calculate node not upgraded and/or slave node
Third upgrade procedure is obtained according to the upgrade command, and is upgraded using the third upgrade procedure.
Optionally, it is described send upgrade command to identified calculate node to be upgraded before, the method can be with
Including：
Switching command is sent to identified calculate node to be upgraded, indicates calculate node to be upgraded by the clothes of self-operating
Business or task switch in the calculate node that upgrading is completed.
In order to achieve the above objectives, the embodiment of the invention also discloses a kind of distributed systems, including：Updating and management equipment,
Host node, slave node and calculate node；Wherein,
The updating and management equipment, for sending upgrade command to slave node；After the completion of detecting slave node upgrading, really
Fixed calculate node to be upgraded；Wherein, identified calculate node quantity to be upgraded is less than the total quantity of calculate node；To determining
Calculate node to be upgraded send upgrade command；Judge whether the quantity for the calculate node that upgrading is completed reaches predetermined threshold value, institute
State the total quantity that predetermined threshold value is not more than calculate node；If not, returning to the step of executing determination calculate node to be upgraded；
If so, the host node is switched to slave node；New host node is determined in the slave node that upgrading is completed；Upgrading is completed
Calculate node be connected to the new host node；The slave node not upgraded into the system and/or the calculating section not upgraded
Point sends upgrade command；
The slave node, for after receiving the upgrade command that the updating and management equipment is sent, carrying out updating operation；
The host node, for after receiving the upgrade command that the updating and management equipment is sent, carrying out updating operation；
The calculate node, for after receiving the upgrade command that the updating and management equipment is sent, carrying out upgrading behaviour
Make.
Optionally, the updating and management equipment is additionally operable to before sending upgrade command to slave node and detects institute
It states in system after the completion of whole node upgradings, switching scheduling strategy；
The host node is additionally operable to, according to current scheduling strategy, adjust itself and save with slave node in the system and calculating
The connection status of point；
The slave node is additionally operable to, according to current scheduling strategy, adjust itself and save with host node in the system and calculating
The connection status of point.
Optionally, the updating and management equipment is additionally operable to after the connection of slave node and calculate node disconnects, will be switched off
The calculate node of connection is connected to host node.
Optionally, the updating and management equipment is additionally operable to after the completion of whole node upgradings in detecting the system, root
According to default concatenate rule, calculate node is connected to corresponding host node or slave node.
Optionally, the updating and management equipment is additionally operable to obtain the first upgrade procedure, first upgrade procedure is sent
To slave node；The second upgrade procedure is obtained, second upgrade procedure is sent to calculate node to be upgraded, so that meter to be upgraded
Operator node is upgraded using second upgrade procedure；Third upgrade procedure is obtained, the third upgrade procedure is sent to
The slave node not upgraded and/or the calculate node not upgraded；
The slave node, is additionally operable to receive the first upgrade procedure that the updating and management equipment is sent, and utilizes described the
One upgrade procedure is upgraded；Alternatively, the first upgrade procedure is obtained according to the upgrade command that the updating and management equipment is sent, and
Upgraded using first upgrade procedure；
The calculate node is additionally operable to receive the second upgrade procedure that the updating and management equipment is sent, and described in utilization
Second upgrade procedure is upgraded；Alternatively, the second upgrade procedure is obtained according to the upgrade command that the updating and management equipment is sent,
And upgraded using second upgrade procedure；
The host node, is additionally operable to receive the third upgrade procedure that the updating and management equipment is sent, and utilizes described the
Three upgrade procedure are upgraded；Alternatively, third upgrade procedure is obtained according to the upgrade command that the updating and management equipment is sent, and
Upgraded using the third upgrade procedure.
Optionally, the calculate node is additionally operable to after receiving the upgrade command that the updating and management equipment is sent, will
The service of self-operating or task switch in the calculate node that upgrading is completed；
Alternatively,
The updating and management equipment is additionally operable to send switching command to identified calculate node to be upgraded；
The calculate node is additionally operable to after receiving the switching command, and the service of self-operating or task are cut
It shifts in the calculate node that upgrading is completed.
Optionally, the calculate node, is additionally operable to after being disconnected with slave node, is connected to host node.
In order to achieve the above objectives, the embodiment of the invention also discloses a kind of updating and management equipment, including：
First sending module, for sending upgrade command to slave node, so that slave node carries out updating operation；
Detection module, for detecting whether slave node upgrades completion, if so, the first determining module of triggering；
First determining module, for determining calculate node to be upgraded；Wherein, identified calculate node quantity to be upgraded is small
In the total quantity of calculate node；
Second sending module, for sending upgrade command to identified calculate node to be upgraded, so that calculating to be upgraded
Node carries out updating operation；
Judgment module, for judging whether the quantity of the calculate node of upgrading completion reaches predetermined threshold value, the default threshold
Total quantity of the value no more than calculate node；If not, triggering first determining module；If so, triggering handover module；
Handover module, for the host node to be switched to slave node；
Second determining module, for determining new host node in the slave node that upgrading is completed；
First link block, the calculate node for completing upgrading are connected to the new host node；
Third sending module, slave node for not upgrading into the system and/or the calculate node not upgraded are sent
Upgrade command, so that the calculate node not upgraded and/or slave node carry out updating operation.
Optionally, the equipment can also include：
Module is disconnected, for before the transmission upgrade command to slave node, passing through and switching scheduling strategy, disconnect
The connection of the connection of slave node and host node and slave node and calculate node；
Link block is established, for after the completion of whole node upgradings in detecting the system, plan to be dispatched by switching
Slightly, the connection of slave node and host node and the connection of slave node and calculate node are established.
Optionally, the equipment can also include：
Second link block, for after disconnecting the connection of slave node and calculate node, will be switched off the calculating section of connection
Point is connected to host node.
Optionally, the equipment can also include：
Third link block, for after the completion of whole node upgradings in detecting the system, being advised according to default connection
Then, calculate node is connected to corresponding host node or slave node.
Optionally, first sending module, specifically can be used for：
The first upgrade procedure is obtained, first upgrade procedure is sent to slave node, so that slave node utilizes described the
One upgrade procedure is upgraded；
Alternatively, upgrade command is sent to slave node, so that slave node obtains the first upgrade procedure according to the upgrade command,
And upgraded using first upgrade procedure；
Second sending module, specifically can be used for：
The second upgrade procedure is obtained, second upgrade procedure is sent to calculate node to be upgraded, so that meter to be upgraded
Operator node is upgraded using second upgrade procedure；
Alternatively, upgrade command is sent to calculate node to be upgraded, so that calculate node to be upgraded is according to the upgrade command
The second upgrade procedure is obtained, and is upgraded using second upgrade procedure；
The third sending module, specifically can be used for：
Third upgrade procedure is obtained, the third upgrade procedure is sent to the slave node not upgraded and/or is not upgraded
Calculate node, so that the calculate node not upgraded and/or slave node are upgraded using the third upgrade procedure；
Alternatively, upgrade command is sent to calculate node to be upgraded, so that the calculate node not upgraded and/or slave node
Third upgrade procedure is obtained according to the upgrade command, and is upgraded using the third upgrade procedure.
Optionally, the equipment can also include：
4th sending module indicates calculating to be upgraded for sending switching command to identified calculate node to be upgraded
Node switches to the service of self-operating or task in the calculate node that upgrading is completed.
It can be seen that in the present solution, upgrade in batches to calculate node, that is, every time to part calculate node into
Row upgrading, other calculate nodes can continue to offer service, in this way, system can continue to offer service while upgrading.
Certainly, implement any of the products of the present invention or method it is not absolutely required at the same reach all the above excellent
Point.
Description of the drawings
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
Obtain other attached drawings according to these attached drawings.
Fig. 1 is a kind of flow diagram of distributed system upgrade method provided in an embodiment of the present invention；
Fig. 2 is a kind of structural schematic diagram of distributed system provided in an embodiment of the present invention；
Fig. 3 is a kind of structural schematic diagram of updating and management equipment provided in an embodiment of the present invention.
Specific implementation mode
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation describes, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
In order to solve the above-mentioned technical problem, an embodiment of the present invention provides a kind of distributed system upgrade method, distributions
System and updating and management equipment, this method can be applied to updating and management equipment, which can be in the system
Equipment, or independently of the equipment of the system, do not limit specifically.
If the updating and management equipment is the equipment in the system, which includes：Host node, calculates section at slave node
Point and updating and management equipment.If the updating and management equipment is the equipment independently of the system, which includes：Host node,
Slave node and calculate node.Alternatively, this method can also be applied to other equipment, do not limit specifically.
Distributed system upgrade method provided in an embodiment of the present invention is described in detail first below.
Fig. 1 is a kind of flow diagram of distributed system upgrade method provided in an embodiment of the present invention, including：
S101：Upgrade command is sent to slave node, so that slave node carries out updating operation.
As an implementation, can first pass through switching scheduling strategy, disconnect slave node and host node connection and
The connection of slave node and calculate node, then execute S101.
For example, it can be scheduled by weight distribution mode, for example, scheduling strategy is " 100 (masters)：0 (standby) "
When, it indicates only to start host node, does not start slave node；Scheduling strategy is " 50 (masters)：When 50 (standby) ", active and standby node is indicated all
Start；Scheduling strategy is " 0 (master)：When 100 (standby) ", indicate only to start slave node, do not start host node ... and specifically dispatch
Strategy can be other, not limit specifically.
In the above example, under normal operating conditions, scheduling strategy can be " 50 (masters)：50 (standby) ", when system needs
When upgrading, scheduling strategy can be switched to " 100 (masters)：0 (standby) " services in this way, being equivalent to stopped in slave node
Operation, specifically, exactly disconnects the connection of slave node and host node and the connection of slave node and calculate node.
As an implementation, after disconnecting the connection of slave node and calculate node, it will be switched off the calculating section of connection
Point is connected to host node.Host node can be managed these calculate nodes instead of slave node, in this way, these calculate nodes are just
Offer service can be provided.
Alternatively, these calculate nodes for being disconnected with slave node, can also be automatically connected to host node, this is also reasonable
's.
After stopping the operation being serviced in slave node, upgrade command is sent to slave node.
The escalation process of node can be the process that new procedures replace original program, or may be that New Parent is replaced
The process of members.For the convenience of description, new procedures or New Parent are referred to as upgrade procedure in the following contents.
In the present embodiment, the equipment for executing this programme can obtain upgrade procedure, and upgrade procedure is sent to node, section
After point receives upgrade procedure, upgraded using upgrade procedure；
Alternatively, the equipment for executing this programme need not obtain upgrade procedure, upgrade command only is sent to node, node connects
After receiving the upgrade command, upgrade procedure is downloaded to preset address, is upgraded using the upgrade procedure of download；
Alternatively, can carry download address in the upgrade command that the equipment for executing this programme is sent, node receives upgrading
After instruction, upgrade procedure is downloaded according to the download address carried in upgrade command, is upgraded using the upgrade procedure of download.Or
Person can also use other schemes, numerous to list herein.
As an implementation, the equipment for executing this programme can obtain the first upgrade procedure, and described first is upgraded
Program is sent to slave node, so that slave node is upgraded using first upgrade procedure.
As another embodiment, the equipment for executing this programme can send upgrade command to slave node, so that standby section
Point obtains the first upgrade procedure according to the upgrade command, and is upgraded using first upgrade procedure.
S102：Whether detection slave node upgrades completion, if so, executing S103.
S103：Determine calculate node to be upgraded；Wherein, identified calculate node quantity to be upgraded is less than calculate node
Total quantity.
In the present embodiment, there are more calculate nodes in distributed system；Slave node can have more, can also only have one
Platform, host node can have more, can also only have one, not limit specifically.
The calculate node quantity to be upgraded determined every time is less than the total quantity of calculate node, that is, is calculating section in batches
Calculate node to be upgraded is determined in point.The calculate node quantity to be upgraded determined every time can be fixed, and can not also fix.
For example, there are 10 calculate nodes in distributed system, can determine 2 calculate nodes to be upgraded every time,
3,3,4 calculate nodes to be upgraded can be determined successively；Determine that the rule of calculate node to be upgraded does not limit.
S104：Upgrade command is sent to identified calculate node to be upgraded, so that calculate node to be upgraded is upgraded
Operation.
As an implementation, the equipment for executing this programme can obtain the second upgrade procedure, and described second is upgraded
Program is sent to calculate node to be upgraded, so that calculate node to be upgraded is upgraded using second upgrade procedure.
As another embodiment, the equipment for executing this programme can send upgrade command to calculate node to be upgraded,
So that calculate node to be upgraded obtains the second upgrade procedure according to the upgrade command, and carried out using second upgrade procedure
Upgrading.
In the present embodiment, the equipment for executing this programme can obtain upgrade procedure, and upgrade procedure is sent to node, section
After point receives upgrade procedure, upgraded using upgrade procedure；
Alternatively, the equipment for executing this programme need not obtain upgrade procedure, upgrade command only is sent to node, node connects
After receiving the upgrade command, upgrade procedure is downloaded to preset address, is upgraded using the upgrade procedure of download；
Alternatively, can carry download address in the upgrade command that the equipment for executing this programme is sent, node receives upgrading
After instruction, upgrade procedure is downloaded according to the download address carried in upgrade command, is upgraded using the upgrade procedure of download.Or
Person can also use other schemes, numerous to list herein.
As an implementation, before sending upgrade command to identified calculate node to be upgraded, can first to
Identified calculate node to be upgraded sends switching command.Calculate node can be after receiving switching command, by self-operating
Service or task switch to other it is promoted completion or this time without in the calculate node of upgrading.
It is appreciated that the service of self-operating or task before upgrading, can be switched to it by calculate node to be upgraded
His promoted completion or this time without in the calculate node of upgrading, in this way, the service being switched or task still can be with
It is run in other calculate nodes, improves the service quality that distributed system externally provides.
Alternatively, as another embodiment, the equipment for executing this programme can not also send switching command, only send and rise
Grade instructs, and in this embodiment, calculate node first cuts the service of self-operating or task after receiving upgrade command
It shifts to other promoted completions or this time without in the calculate node of upgrading, then carries out updating operation, this is also reasonable
's.
S105：Judge whether the quantity for the calculate node that upgrading is completed reaches predetermined threshold value, is executed if not, returning
S103, if so, executing S106.
Predetermined threshold value in S105 is less than or equal to the total quantity of calculate node in distributed system, for example, predetermined threshold value
Can be the half of the total quantity, or 2/3rds of the total quantity, or a quarter of the total quantity, etc.
Deng being not limited thereto.
S106：The host node is switched to slave node.
It will be understood by those skilled in the art that in a distributed system, host node and slave node are all to calculate node
It is scheduled management, host node can switch mutually with slave node.
S107：New host node is determined in the slave node that upgrading is completed.
S108：The calculate node that upgrading is completed is connected to the new host node.
In an alternate embodiment of the present invention where, if applying the embodiment of above-mentioned " switching scheduling strategy ", this
In, after former host node is switched to slave node, former host node is still applied to the scheduling strategy, that is, former host node and other sections
The connection of point disconnects.And new host node is also applied for the scheduling strategy, that is, establish new host node and calculate node
Connection.In this way, the calculate node that upgrading is completed can be connected to the new host node.
S109：The slave node not upgraded into the system and/or the calculate node not upgraded send upgrade command, so that
The calculate node not upgraded and/or slave node carry out updating operation.
As an implementation, the equipment for executing this programme can obtain third upgrade procedure, and the third is upgraded
Program is sent to the slave node not upgraded and/or the calculate node not upgraded, so that the calculate node not upgraded and/or standby
Node is upgraded using the third upgrade procedure.
As another embodiment, the equipment for executing this programme can send upgrade command to calculate node to be upgraded,
So that the calculate node not upgraded and/or slave node obtain third upgrade procedure according to the upgrade command, and utilize institute
Third upgrade procedure is stated to be upgraded.
In the present embodiment, the equipment for executing this programme can obtain upgrade procedure, and upgrade procedure is sent to node, section
After point receives upgrade procedure, upgraded using upgrade procedure；
Alternatively, the equipment for executing this programme need not obtain upgrade procedure, upgrade command only is sent to node, node connects
After receiving the upgrade command, upgrade procedure is downloaded to preset address, is upgraded using the upgrade procedure of download；
Alternatively, can carry download address in the upgrade command that the equipment for executing this programme is sent, node receives upgrading
After instruction, upgrade procedure is downloaded according to the download address carried in upgrade command, is upgraded using the upgrade procedure of download.Or
Person can also use other schemes, numerous to list herein.
As an implementation, it after the completion of whole node upgradings in detecting the system, can be adjusted by switching
Degree strategy, establishes the connection of slave node and host node and the connection of slave node and calculate node.
It is emphasized that " establishing connection " here refers to starting slave node, allow slave node and other node phases
Connection, and be not specific to establish the connection of slave node and some node.
Continue the example of above-mentioned switchover policy, that is, by scheduling strategy by " 100 (masters)：0 (standby) " is switched to " 50
(master)：50 (standby) ", in this way, active and standby node is all run.
It as an implementation, can be after the completion of whole node upgradings in detecting the system, according to default company
Rule is connect, calculate node is connected to corresponding host node or slave node.
The default concatenate rule can be pre-stored host node and calculate node correspondence and slave node with
Calculate node is connected to its corresponding host node or slave node, host node and slave node pair by the correspondence of calculate node
The calculate node of itself connection is managed.Or, or other, are for example, default concatenate rule is that " host node corresponds to 10
Platform calculate node, slave node correspond to 5 calculate nodes ", in this way, can be randomly assigned specific corresponding to relationship.In the present embodiment,
Specifically default concatenate rule does not limit.
Alternatively, calculate node can also be automatically connected to host node or slave node, this is also rational.
According to previously mentioned, total quantity of the above-mentioned predetermined threshold value no more than calculate node.
As an implementation, predetermined threshold value can be equal to the total quantity of calculate node, that is to say, that present embodiment
In, after the completion of all calculate nodes all upgrade, then execute S106-S109.
In present embodiment, after the completion of all calculate nodes all upgrade, host node is switched to slave node, is completed in upgrading
Slave node in determine new host node, all calculate nodes are connected to the new host node, what is do not upgraded into system is standby
Node (namely former host node) sends upgrade command, and after the completion of former host node upgrading, whole nodes all upgrade completion in system.
As an example it is assumed that having a host node A, a slave node B, 6 calculate nodes in system：Calculate node C, meter
Operator node D, calculate node E, calculate node F, calculate node G, calculate node H.Assuming that host node A and calculate node C, D, E phase
Connection, and it is managed, slave node B is connected with calculate node F, G, H, and is managed to it.
1, by switching scheduling strategy, disconnect slave node B and host node A connection and slave node B and calculate node F,
G, the connection of H.
2, calculate node F, G, H are connected to host node A, host node A is managed calculate node F, G, H.
3, upgrade command is sent to slave node B, slave node B carries out updating operation.Specific escalation process previous contents are
Through describing, which is not described herein again.
4, after the completion of slave node B upgradings, in above-mentioned 6 calculate nodes, calculate node to be upgraded is determined.In order to which simplification is retouched
It states, only includes a slave node in this example, if in system including more slave nodes, can detect all for section
After the completion of point all upgrades, calculate node to be upgraded is determined.
5, assume to determine that the rule of calculate node to be upgraded is " determining 3 every time ", this time determine that calculate node D, E, F are
Calculate node to be upgraded.
6, upgrade command is sent to calculate node D, E, F to be upgraded, calculate node D, E, F carry out updating operation.
7, judge whether the quantity for the calculate node that upgrading is completed reaches predetermined threshold value.It is assumed that predetermined threshold value is equal to meter
The total quantity 6 of operator node, at this moment, the quantity for upgrading the calculate node of completion is 3, not up to 6, and judging result is no.
8, continue to determine calculate node to be upgraded.At this point, having existed the calculate node that upgrading is completed；In the meter not upgraded
In operator node, calculate node to be upgraded is determined.This time determine that calculate node C, G, H are calculate node to be upgraded.
9, upgrade command is sent to calculate node C, G, H to be upgraded, calculate node C, G, H carry out updating operation.
10, judge whether the quantity for the calculate node that upgrading is completed reaches predetermined threshold value.At this moment, upgrade the calculating section of completion
The quantity of point is 6, reaches predetermined threshold value, judging result is yes.
11, host node A is switched to slave node.
12, the slave node B that upgrading is completed is determined as new host node.To simplify the description, only include in this example
One slave node can upgrade at more if including more slave nodes in system and determine new master in the slave node completed
Node.
13,6 calculate nodes that upgrading is completed are connected to new host node, that is, former slave node B.
In an alternate embodiment of the present invention where, after former host node A being switched to slave node, former host node A is still applicable in
It is disconnected in the scheduling strategy, that is, the connection of original host node A and slave node and the connection of original host node A and calculate node.
And new host node B is also applied for the scheduling strategy, that is, establish the connection of new host node B and calculate node.In this way, rising
The calculate node that grade is completed can be connected to new host node B.
14, the slave node not upgraded into system, that is, former host node A send upgrade command, and former host node A is risen
Grade operation.In this way, whole node upgradings are completed in system.
15, by switching scheduling strategy, the connection of host node and slave node and the company of slave node and calculate node are established
It connects.
16, according to concatenate rule is preset, calculate node is connected to corresponding host node or slave node.
Assuming that default concatenate rule is that " node A corresponds to calculate node C, D, E；Node B corresponds to calculate node E, F, G ", then
Calculate node C, D, E are reconnected into node A.
Optionally, node A can also be switched to host node again, can not also switches, does not limit specifically.
Using the above embodiment, calculate node is upgraded in batches, that is, every time to part calculate node into
Row upgrading, other calculate nodes can continue to offer service, in this way, system can continue to offer service while upgrading.Into
One step, it is promoted can be switched to other before upgrading by calculate node to be upgraded for the service of self-operating or task
It completes or this time without in the calculate node of upgrading, in this way, distributed system can be made not influenced by upgrading, continuing
Service is externally provided, the service quality that distributed system externally provides is improved.
As another embodiment, predetermined threshold value can be less than the total quantity of calculate node, that is to say, that this embodiment party
In formula, after the completion of the upgrading of part calculate node, S106-S109 is just executed.For example, predetermined threshold value can be calculate node total quantity
Half, in this case, half calculate node upgrading after the completion of, just execute S106-S109.
Specifically, after the completion of the upgrading of half calculate node, host node is switched to slave node, in the slave node that upgrading is completed
The half calculate node that upgrading is completed is connected to the new host node, not upgraded into system by the new host node of middle determination
Slave node (namely former host node) and the other half calculate node not upgraded send upgrade command, former host node and should
After the completion of the upgrading of the other half calculate node, whole nodes all upgrade completion in system.
As an example it is assumed that having a host node A, a slave node B, 6 calculate nodes in system：Calculate node C, meter
Operator node D, calculate node E, calculate node F, calculate node G, calculate node H.Assuming that host node A and calculate node C, D, E phase
Connection, and it is managed, slave node B is connected with calculate node F, G, H, and is managed to it.
1, by switching scheduling strategy, disconnect slave node B and host node A connection and slave node B and calculate node F,
G, the connection of H.
2, calculate node F, G, H are connected to host node A, host node A is managed calculate node F, G, H.
3, upgrade command is sent to slave node B, slave node B carries out updating operation.Specific escalation process previous contents are
Through describing, which is not described herein again.
4, after the completion of slave node B upgradings, in above-mentioned 6 calculate nodes, calculate node to be upgraded is determined.In order to which simplification is retouched
It states, only includes a slave node in this example, if in system including more slave nodes, can detect all for section
After the completion of point all upgrades, calculate node to be upgraded is determined.
5, assume to determine that the rule of calculate node to be upgraded is " determining 3 every time ", this time determine that calculate node D, E, F are
Calculate node to be upgraded.
6, upgrade command is sent to calculate node D, E, F to be upgraded, calculate node D, E, F carry out updating operation.
7, judge whether the quantity for the calculate node that upgrading is completed reaches predetermined threshold value.It is calculate node to here preset at threshold value
The half of total quantity, that is, 3, at this moment, the quantity for upgrading the calculate node of completion is 3, and judging result is yes.
8, host node A is switched to slave node.
9, the slave node B that upgrading is completed is determined as new host node.To simplify the description, only include one in this example
Platform slave node can upgrade at more if including more slave nodes in system and determine new main section in the slave node completed
Point.
10,3 calculate nodes D, E, F that upgrading is completed are connected to new host node, that is, former slave node B.
In an alternate embodiment of the present invention where, after former host node A being switched to slave node, former host node A is still applicable in
It is disconnected in the scheduling strategy, that is, the connection of original host node A and slave node and the connection of original host node A and calculate node.
And new host node B is also applied for the scheduling strategy, that is, establish the connection of new host node B and calculate node.In this way, rising
Calculate node D, E, F that grade is completed can be connected to new host node B.
The slave node (namely original host node A) 11, not upgraded into system and the 3 calculate node C not upgraded,
G, H sends upgrade command, and former host node A and calculate node C, G, H carry out updating operation.In this way, whole node liters in system
Grade is completed.
12, by switching scheduling strategy, the connection of host node and slave node and the company of slave node and calculate node are established
It connects.
13, according to concatenate rule is preset, calculate node is connected to corresponding host node or slave node.
Assuming that default concatenate rule is that " node A corresponds to calculate node C, D, E；Node B corresponds to calculate node E, F, G ", then
Calculate node C, D, E are connected to node A, calculate node E, F, G are connected to node B.
Optionally, node A can also be switched to host node again, can not also switches, does not limit specifically.
In an alternate embodiment of the present invention where, " predetermined threshold value is the half of calculate node total quantity " is only this implementation
A kind of concrete condition that example provides does not constitute " predetermined threshold value " and limits.
In above-mentioned escalation process, calculate node D, E, F for upgrading completion are connected with the node B that upgrading is completed, and do not rise
Calculate node C, G, H of grade are connected with the node A not upgraded；That is, there is no upgrade the node completed and do not upgrade
Node connection；Therefore, this programme can also solve the problems, such as that system upgrade new and old edition is incompatible.
It is appreciated that the active/standby node that the calculate node that upgrading is completed is completed with upgrading is calculating that is compatible, not upgrading
Node is compatible with the active/standby node not upgraded；And in above-mentioned escalation process, upgrade calculate node D, E, F and liter of completion
The node B that grade is completed is connected, and calculate node C, G, the H not upgraded are connected with the node A not upgraded, and new and old edition is not present
Incompatible problem.
Using the above embodiment, in a first aspect, upgrading in batches to calculate node, that is, part is counted every time
Operator node is upgraded, and other calculate nodes can continue to offer service, in this way, system can continue to carry while upgrading
For service.The service of self-operating or task can be switched to other by second aspect, calculate node to be upgraded before upgrading
Promoted completion or this time without in the calculate node of upgrading, in this way, can make distributed system not by upgrading shadow
It rings, external offer service is provided, improve the service quality that distributed system externally provides.The third aspect, before solving upgrading
The system problem incompatible with the system after upgrading.
Corresponding with above method embodiment, the embodiment of the present invention also provides a kind of distributed system.
Fig. 2 is a kind of structural schematic diagram of distributed system provided in an embodiment of the present invention, including：Updating and management equipment,
Host node, slave node and calculate node.In the present embodiment, there are more calculate nodes (calculate node 1, meter in distributed system
Operator node 2 ... calculate node N)；Slave node can have more, can also only have one, host node can have more, can also
It only one, does not limit specifically.
In the embodiment depicted in figure 2, updating and management equipment, for sending upgrade command to slave node；Detecting standby section
After the completion of point upgrading, calculate node to be upgraded is determined；Wherein, identified calculate node quantity to be upgraded is less than calculate node
Total quantity；Upgrade command is sent to identified calculate node to be upgraded；Judge upgrading complete calculate node quantity whether
Reach predetermined threshold value, the predetermined threshold value is not more than the total quantity of calculate node；If not, it is to be upgraded to return to the execution determination
The step of calculate node；If so, the host node is switched to slave node；New master is determined in the slave node that upgrading is completed
Node；The calculate node that upgrading is completed is connected to the new host node；The slave node that does not upgrade into the system and/or
The calculate node not upgraded sends upgrade command；
Slave node, for after receiving the upgrade command that the updating and management equipment is sent, carrying out updating operation；
Host node, for after receiving the upgrade command that the updating and management equipment is sent, carrying out updating operation；
Calculate node, for after receiving the upgrade command that the updating and management equipment is sent, carrying out updating operation.
As an implementation, updating and management equipment, can be also used for slave node send upgrade command before, with
And it detects in the system after the completion of whole node upgradings, switching scheduling strategy；
Host node can be also used for, according to current scheduling strategy, adjusting itself and saving with slave node in the system and calculating
The connection status of point；
Slave node, is additionally operable to according to current scheduling strategy, adjusts itself with host node in the system and calculate node
Connection status.
For example, it can be scheduled by weight distribution mode, for example, scheduling strategy is " 100 (masters)：0 (standby) "
When, it indicates only to start host node, does not start slave node；Scheduling strategy is " 50 (masters)：When 50 (standby) ", active and standby node is indicated all
Start；Scheduling strategy is " 0 (master)：When 100 (standby) ", indicate only to start slave node, do not start host node ... and specifically dispatch
Strategy can be other, not limit specifically.
Under normal operating conditions, scheduling strategy can be " 50 (masters)：50 (standby) ", when system needs upgrading, upgrading pipe
Scheduling strategy can be switched to " 100 (masters) by reason equipment：0 (standby) ", in this way, being equivalent to stopped the fortune serviced in slave node
Row, specifically, exactly disconnects the connection of slave node and host node and the connection of slave node and calculate node.
It, can be by scheduling strategy by " 100 (masters) after the completion of whole node upgradings in detecting the system：0 (standby) "
It is switched to " 50 (masters)：50 (standby) ", in this way, active and standby node is all run.
Alternatively, updating and management equipment can only switch scheduling strategy, the operation of connection is disconnected/established by active/standby section
Point executes.For example, scheduling strategy is switched to " 100 (masters) by updating and management equipment：After 0 (standby) ", slave node disconnects automatically and master
The connection of node and calculate node；Scheduling strategy is switched to " 50 (masters) by updating and management equipment：After 50 (standby) ", slave node is automatic
Establish the connection with host node and calculate node.
As an implementation, updating and management equipment can be also used for disconnecting in the connection of slave node and calculate node
Later, the calculate node that will be switched off connection is connected to host node.
As an implementation, updating and management equipment can be also used for whole node liters in detecting the system
After the completion of grade, according to default concatenate rule, calculate node is connected to corresponding host node or slave node.
As an implementation, updating and management equipment can be also used for obtaining the first upgrade procedure, by described first liter
Grade program is sent to slave node；The second upgrade procedure is obtained, second upgrade procedure is sent to calculate node to be upgraded, with
Calculate node to be upgraded is set to be upgraded using second upgrade procedure；Third upgrade procedure is obtained, the third is upgraded
Program is sent to the slave node not upgraded and/or the calculate node not upgraded；
The slave node, is additionally operable to receive the first upgrade procedure that the updating and management equipment is sent, and utilizes described the
One upgrade procedure is upgraded；Alternatively, the first upgrade procedure is obtained according to the upgrade command that the updating and management equipment is sent, and
Upgraded using first upgrade procedure；
The calculate node is additionally operable to receive the second upgrade procedure that the updating and management equipment is sent, and described in utilization
Second upgrade procedure is upgraded；Alternatively, the second upgrade procedure is obtained according to the upgrade command that the updating and management equipment is sent,
And upgraded using second upgrade procedure；
The host node, is additionally operable to receive the third upgrade procedure that the updating and management equipment is sent, and utilizes described the
Three upgrade procedure are upgraded；Alternatively, third upgrade procedure is obtained according to the upgrade command that the updating and management equipment is sent, and
Upgraded using the third upgrade procedure.
As an implementation, the updating and management equipment is additionally operable to send to identified calculate node to be upgraded
Switching command；
The calculate node is additionally operable to after receiving the switching command, and the service of self-operating or task are cut
Shift to other promoted completions or this time without in the calculate node of upgrading.
Alternatively, as another embodiment, the calculate node can be also used for setting receiving the updating and management
After the upgrade command that preparation is sent, by the service of self-operating or task switch to other it is promoted completion or this time not
In the calculate node upgraded.
The calculate node, can be also used for after being disconnected with slave node, be connected to host node；
That is, the operation that connection is disconnected, established between node can be executed by updating and management equipment, it can also
It is executed automatically by node.Similar, the operation that service/task of calculate node self-operating switches to other calculate nodes can be with
It is executed, can also be automatically executed by calculate node by updating and management equipment.
Using embodiment illustrated in fig. 2 of the present invention, calculate node is upgraded in batches, that is, part is calculated every time
Node is upgraded, and other calculate nodes can continue to offer service, in this way, system can continue to provide while upgrading
Service.Further, the service of self-operating or task can have been switched to other by calculate node to be upgraded before upgrading
Promoted completion or this time without in the calculate node of upgrading, in this way, can make distributed system not by upgrading shadow
It rings, external offer service is provided, improve the service quality that distributed system externally provides.
Corresponding with above method embodiment, the embodiment of the present invention also provides a kind of updating and management equipment.
Fig. 3 is a kind of structural schematic diagram of updating and management equipment provided in an embodiment of the present invention, including：
First sending module 301, for sending upgrade command to slave node, so that slave node carries out updating operation；
Detection module 302, for detecting whether slave node upgrades completion, if so, the first determining module 303 of triggering；
First determining module 303, for determining calculate node to be upgraded；Wherein, identified calculate node number to be upgraded
Total quantity of the amount less than calculate node；
Second sending module 304, for sending upgrade command to identified calculate node to be upgraded, so that meter to be upgraded
Operator node carries out updating operation；
Judgment module 305, it is described default for judging whether the quantity of the calculate node of upgrading completion reaches predetermined threshold value
Threshold value is not more than the total quantity of calculate node；If not, triggering first determining module 303；If so, triggering handover module
306；
Handover module 306, for the host node to be switched to slave node；
Second determining module 307, for determining new host node in the slave node that upgrading is completed；
First link block 308, the calculate node for completing upgrading are connected to the new host node；
Third sending module 309, slave node for not upgrading into the system and/or the calculate node hair not upgraded
Upgrade command is sent, so that the calculate node not upgraded and/or slave node carry out updating operation.
As an implementation, the equipment can also include：It disconnects module and establishes link block (in figure not
Show), wherein
Module is disconnected, for before the transmission upgrade command to slave node, passing through and switching scheduling strategy, disconnect
The connection of the connection of slave node and host node and slave node and calculate node；
Link block is established, for after the completion of whole node upgradings in detecting the system, plan to be dispatched by switching
Slightly, the connection of slave node and host node and the connection of slave node and calculate node are established.
As an implementation, the equipment can also include：
Second link block (not shown), for after disconnecting the connection of slave node and calculate node, will be switched off
The calculate node of connection is connected to host node.
As an implementation, the equipment can also include：
Third link block (not shown), for after the completion of whole node upgrading in detecting the system, root
According to default concatenate rule, calculate node is connected to corresponding host node or slave node.
As an implementation, the first sending module 301, specifically can be used for：
The first upgrade procedure is obtained, first upgrade procedure is sent to slave node, so that slave node utilizes described the
One upgrade procedure is upgraded；
Alternatively, upgrade command is sent to slave node, so that slave node obtains the first upgrade procedure according to the upgrade command,
And upgraded using first upgrade procedure；
Second sending module 304, specifically can be used for：
The second upgrade procedure is obtained, second upgrade procedure is sent to calculate node to be upgraded, so that meter to be upgraded
Operator node is upgraded using second upgrade procedure；
Alternatively, upgrade command is sent to calculate node to be upgraded, so that calculate node to be upgraded is according to the upgrade command
The second upgrade procedure is obtained, and is upgraded using second upgrade procedure；
Third sending module 309, specifically can be used for：
Third upgrade procedure is obtained, the third upgrade procedure is sent to the slave node not upgraded and/or is not upgraded
Calculate node, so that the calculate node not upgraded and/or slave node are upgraded using the third upgrade procedure；
Alternatively, upgrade command is sent to calculate node to be upgraded, so that the calculate node not upgraded and/or slave node
Third upgrade procedure is obtained according to the upgrade command, and is upgraded using the third upgrade procedure.
As an implementation, the equipment can also include：
4th sending module (not shown) refers to for sending switching command to identified calculate node to be upgraded
Show that calculate node to be upgraded switches to the service of self-operating or task in the calculate node that upgrading is completed.
Using embodiment illustrated in fig. 3 of the present invention, calculate node is upgraded in batches, that is, part is calculated every time
Node is upgraded, and other calculate nodes can continue to offer service, in this way, system can continue to provide while upgrading
Service.Further, the service of self-operating or task can have been switched to other by calculate node to be upgraded before upgrading
Promoted completion or this time without in the calculate node of upgrading, in this way, can make distributed system not by upgrading shadow
It rings, external offer service is provided, improve the service quality that distributed system externally provides.
In an alternate embodiment of the present invention where, herein, relational terms such as first and second and the like are only
Only it is used for distinguishing one entity or operation from another entity or operation, without necessarily requiring or implying these realities
There are any actual relationship or orders between body or operation.Moreover, the terms "include", "comprise" or its it is any its
He is intended to non-exclusive inclusion by variant, so that the process, method, article or equipment including a series of elements is not
Only include those elements, but also include other elements that are not explicitly listed, or further include for this process, method,
Article or the intrinsic element of equipment.In the absence of more restrictions, being wanted by what sentence "including a ..." limited
Element, it is not excluded that there is also other identical elements in the process, method, article or apparatus that includes the element.
Each embodiment in this specification is all made of relevant mode and describes, identical similar portion between each embodiment
Point just to refer each other, and each embodiment focuses on the differences from other embodiments.Especially for system reality
For applying example, since it is substantially similar to the method embodiment, so description is fairly simple, related place is referring to embodiment of the method
Part explanation.
One of ordinary skill in the art will appreciate that realizing that all or part of step of above-described embodiment can pass through hardware
It completes, relevant hardware can also be instructed to complete by program, the program can be stored in a kind of computer-readable
In storage medium, storage medium mentioned above can be read-only memory, disk or CD etc..
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the scope of the present invention.It is all
Any modification, equivalent replacement, improvement and so within the spirit and principles in the present invention, are all contained in protection scope of the present invention
It is interior.
Claims (19)
1. a kind of distributed system upgrade method, which is characterized in that the distributed system includes：Host node, slave node and meter
Operator node；The method includes：
Upgrade command is sent to slave node, so that slave node carries out updating operation；
After the completion of detecting slave node upgrading, calculate node to be upgraded is determined；Wherein, identified calculate node number to be upgraded
Total quantity of the amount less than calculate node；
Upgrade command is sent to identified calculate node to be upgraded, so that calculate node to be upgraded carries out updating operation；
Judge whether the quantity for the calculate node that upgrading is completed reaches predetermined threshold value, the predetermined threshold value is no more than calculate node
Total quantity；
If not, returning to the step of executing determination calculate node to be upgraded；
If so, the host node is switched to slave node；New host node is determined in the slave node that upgrading is completed；It will upgrading
The calculate node of completion is connected to the new host node；The slave node not upgraded into the system and/or the meter not upgraded
Operator node sends upgrade command, so that the calculate node not upgraded and/or slave node carry out updating operation.
2. according to the method described in claim 1, it is characterized in that, before the transmission upgrade command to slave node, also wrap
It includes：
By switching scheduling strategy, slave node and the connection of host node and the connection of slave node and calculate node are disconnected；
The method further includes：
After the completion of whole node upgradings in detecting the system, by switching scheduling strategy, slave node and host node are established
Connection and slave node and calculate node connection.
3. according to the method described in claim 2, it is characterized in that, after disconnecting the connection of slave node and calculate node, go back
Including：
The calculate node that will be switched off connection is connected to host node.
4. method according to claim 1 or 3, which is characterized in that the method further includes：
After the completion of whole node upgrading in detecting the system, according to default concatenate rule, calculate node is connected to pair
The host node or slave node answered.
5. according to the method described in claim 1, it is characterized in that, described send upgrade command to slave node, so that slave node
Updating operation is carried out, including：
The first upgrade procedure is obtained, first upgrade procedure is sent to slave node, so that slave node utilizes described first liter
Grade program is upgraded；
Alternatively, upgrade command is sent to slave node, so that slave node obtains the first upgrade procedure, and profit according to the upgrade command
Upgraded with first upgrade procedure；
It is described to send upgrade command to identified calculate node to be upgraded, so that calculate node to be upgraded carries out updating operation,
Including：
The second upgrade procedure is obtained, second upgrade procedure is sent to calculate node to be upgraded, so that calculating to be upgraded saves
Point is upgraded using second upgrade procedure；
Alternatively, upgrade command is sent to calculate node to be upgraded, so that calculate node to be upgraded is obtained according to the upgrade command
Second upgrade procedure, and upgraded using second upgrade procedure；
The slave node not upgraded into the system and/or the calculate node not upgraded send upgrade command, so that described
The calculate node and/or slave node not upgraded carry out updating operation, including：
Obtain third upgrade procedure, the calculating that the third upgrade procedure is sent to the slave node not upgraded and/or do not upgraded
Node, so that the calculate node not upgraded and/or slave node are upgraded using the third upgrade procedure；
Alternatively, send upgrade command to calculate node to be upgraded so that the calculate node not upgraded and/or slave node according to
The upgrade command obtains third upgrade procedure, and is upgraded using the third upgrade procedure.
6. according to the method described in claim 1, it is characterized in that, being risen in described sent to identified calculate node to be upgraded
Before grade instruction, further include：
Send switching command to identified calculate node to be upgraded, indicate calculate node to be upgraded by the service of self-operating or
Person's task switches in the calculate node that upgrading is completed.
7. a kind of distributed system, which is characterized in that including：Updating and management equipment, host node, slave node and calculate node；Its
In,
The updating and management equipment, for sending upgrade command to slave node；After the completion of detecting slave node upgrading, determination waits for
Upgrade calculate node；Wherein, identified calculate node quantity to be upgraded is less than the total quantity of calculate node；It is waited for identified
Upgrade calculate node and sends upgrade command；Judge whether the quantity for the calculate node that upgrading is completed reaches predetermined threshold value, it is described pre-
If threshold value is not more than the total quantity of calculate node；If not, returning to the step of executing determination calculate node to be upgraded；If
It is that the host node is switched to slave node；New host node is determined in the slave node that upgrading is completed；The meter that upgrading is completed
Operator node is connected to the new host node；The slave node not upgraded into the system and/or the calculate node not upgraded hair
Send upgrade command；
The slave node, for after receiving the upgrade command that the updating and management equipment is sent, carrying out updating operation；
The host node, for after receiving the upgrade command that the updating and management equipment is sent, carrying out updating operation；
The calculate node, for after receiving the upgrade command that the updating and management equipment is sent, carrying out updating operation.
8. system according to claim 7, which is characterized in that
The updating and management equipment is additionally operable to before sending upgrade command to slave node and detects and is complete in the system
After the completion of portion's node upgrading, switch scheduling strategy；
The host node, is additionally operable to according to current scheduling strategy, adjusts itself with slave node in the system and calculate node
Connection status；
The slave node, is additionally operable to according to current scheduling strategy, adjusts itself with host node in the system and calculate node
Connection status.
9. system according to claim 8, which is characterized in that
The updating and management equipment is additionally operable to after the connection of slave node and calculate node disconnects, will be switched off the calculating of connection
Node is connected to host node.
10. the system according to claim 7 or 9, which is characterized in that
The updating and management equipment is additionally operable to after the completion of whole node upgradings in detecting the system, according to default connection
Calculate node is connected to corresponding host node or slave node by rule.
11. system according to claim 7, which is characterized in that
The updating and management equipment is additionally operable to obtain the first upgrade procedure, first upgrade procedure is sent to slave node；It obtains
The second upgrade procedure is obtained, second upgrade procedure is sent to calculate node to be upgraded, so that calculate node to be upgraded utilizes
Second upgrade procedure is upgraded；Obtain third upgrade procedure, the third upgrade procedure is sent to do not upgrade it is standby
Node and/or the calculate node not upgraded；
The slave node is additionally operable to receive the first upgrade procedure that the updating and management equipment is sent, and utilizes described first liter
Grade program is upgraded；Alternatively, obtaining the first upgrade procedure according to the upgrade command that the updating and management equipment is sent, and utilize
First upgrade procedure is upgraded；
The calculate node is additionally operable to receive the second upgrade procedure that the updating and management equipment is sent, and utilizes described second
Upgrade procedure is upgraded；Alternatively, obtaining the second upgrade procedure, and profit according to the upgrade command that the updating and management equipment is sent
Upgraded with second upgrade procedure；
The host node is additionally operable to receive the third upgrade procedure that the updating and management equipment is sent, and utilizes the third liter
Grade program is upgraded；Alternatively, obtaining third upgrade procedure according to the upgrade command that the updating and management equipment is sent, and utilize
The third upgrade procedure is upgraded.
12. system according to claim 7, which is characterized in that
The calculate node is additionally operable to after receiving the upgrade command that the updating and management equipment is sent, by self-operating
Service or task switch in the calculate node that upgrading is completed；
Alternatively,
The updating and management equipment is additionally operable to send switching command to identified calculate node to be upgraded；The calculate node,
It is additionally operable to after receiving the switching command, the service of self-operating or task is switched into the calculate node that upgrading is completed
In.
13. system according to claim 7, which is characterized in that the calculate node is additionally operable to disconnect company with slave node
After connecing, it is connected to host node.
14. a kind of updating and management equipment, which is characterized in that including：
First sending module, for sending upgrade command to slave node, so that slave node carries out updating operation；
Detection module, for detecting whether slave node upgrades completion, if so, the first determining module of triggering；
First determining module, for determining calculate node to be upgraded；Wherein, identified calculate node quantity to be upgraded is less than meter
The total quantity of operator node；
Second sending module, for sending upgrade command to identified calculate node to be upgraded, so that calculate node to be upgraded
Carry out updating operation；
Judgment module, for judging whether the quantity of the calculate node of upgrading completion reaches predetermined threshold value, the predetermined threshold value is not
More than the total quantity of calculate node；If not, triggering first determining module；If so, triggering handover module；
Handover module, for the host node to be switched to slave node；
Second determining module, for determining new host node in the slave node that upgrading is completed；
First link block, the calculate node for completing upgrading are connected to the new host node；
Third sending module, slave node for not upgrading into the system and/or the calculate node not upgraded send upgrading
Instruction, so that the calculate node not upgraded and/or slave node carry out updating operation.
15. equipment according to claim 14, which is characterized in that the equipment further includes：
Module is disconnected, for before the transmission upgrade command to slave node, by switching scheduling strategy, disconnecting standby section
The connection of connection and slave node and calculate node of the point with host node；
Link block is established, for after the completion of whole node upgradings in detecting the system, by switching scheduling strategy, building
Vertical slave node and the connection of host node and the connection of slave node and calculate node.
16. equipment according to claim 15, which is characterized in that the equipment further includes：
Second link block, for after disconnecting the connection of slave node and calculate node, the calculate node that will be switched off connection to connect
It is connected to host node.
17. the equipment according to claim 14 or 16, which is characterized in that the equipment further includes：
Third link block, for after the completion of whole node upgradings in detecting the system, according to default concatenate rule, inciting somebody to action
Calculate node is connected to corresponding host node or slave node.
18. equipment according to claim 14, which is characterized in that first sending module is specifically used for：
The first upgrade procedure is obtained, first upgrade procedure is sent to slave node, so that slave node utilizes described first liter
Grade program is upgraded；
Alternatively, upgrade command is sent to slave node, so that slave node obtains the first upgrade procedure, and profit according to the upgrade command
Upgraded with first upgrade procedure；
Second sending module, is specifically used for：
The second upgrade procedure is obtained, second upgrade procedure is sent to calculate node to be upgraded, so that calculating to be upgraded saves
Point is upgraded using second upgrade procedure；
Alternatively, upgrade command is sent to calculate node to be upgraded, so that calculate node to be upgraded is obtained according to the upgrade command
Second upgrade procedure, and upgraded using second upgrade procedure；
The third sending module, is specifically used for：
Obtain third upgrade procedure, the calculating that the third upgrade procedure is sent to the slave node not upgraded and/or do not upgraded
Node, so that the calculate node not upgraded and/or slave node are upgraded using the third upgrade procedure；
Alternatively, send upgrade command to calculate node to be upgraded so that the calculate node not upgraded and/or slave node according to
The upgrade command obtains third upgrade procedure, and is upgraded using the third upgrade procedure.
19. equipment according to claim 14, which is characterized in that the equipment further includes：
4th sending module indicates calculate node to be upgraded for sending switching command to identified calculate node to be upgraded
The service of self-operating or task are switched in the calculate node that upgrading is completed.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710200566.2A CN108667641A (en) | 2017-03-30 | 2017-03-30 | Distributed system upgrade method, distributed system and updating and management equipment |
| PCT/CN2018/081357 WO2018177417A1 (en) | 2017-03-30 | 2018-03-30 | Method of upgrading distributed system, distributed system, and upgrade management device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710200566.2A CN108667641A (en) | 2017-03-30 | 2017-03-30 | Distributed system upgrade method, distributed system and updating and management equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108667641A true CN108667641A (en) | 2018-10-16 |
Family
ID=63675257
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710200566.2A Pending CN108667641A (en) | 2017-03-30 | 2017-03-30 | Distributed system upgrade method, distributed system and updating and management equipment |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN108667641A (en) |
| WO (1) | WO2018177417A1 (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110247791A (en) * | 2019-05-22 | 2019-09-17 | 北京英视睿达科技有限公司 | The environmental monitoring system and update method of work online updating free of discontinuities |
| CN110597531A (en) * | 2019-08-05 | 2019-12-20 | 平安科技（深圳）有限公司 | Distributed module upgrading method and device and storage medium |
| CN111111161A (en) * | 2019-12-19 | 2020-05-08 | 上海米哈游天命科技有限公司 | Game server repairing method, device, server and medium |
| CN111930448A (en) * | 2020-09-23 | 2020-11-13 | 南京梦饷网络科技有限公司 | Method, electronic device, and storage medium for service distribution |
| CN114201549A (en) * | 2020-09-17 | 2022-03-18 | 金篆信科有限责任公司 | Switching method, system, server and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101217678A (en) * | 2007-12-28 | 2008-07-09 | 华为技术有限公司 | System upgrade method, system and device |
| CN101741894A (en) * | 2008-11-26 | 2010-06-16 | ***通信集团公司 | Upgrade method for distributed system and upgrade scheduling node and system |
| US20140188955A1 (en) * | 2001-06-05 | 2014-07-03 | Silicon Graphics International Corp. | Clustered filesystem with membership version support |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7143406B2 (en) * | 2001-01-23 | 2006-11-28 | Wildtangent, Inc. | Asynchronous software update |
| US9170852B2 (en) * | 2012-02-02 | 2015-10-27 | Microsoft Technology Licensing, Llc | Self-updating functionality in a distributed system |
| US10042626B2 (en) * | 2015-06-29 | 2018-08-07 | Verizon Patent And Licensing Inc. | Software updates using client self-reporting and a hierarchical data structure |
| CN106354531B (en) * | 2016-08-25 | 2020-03-27 | 杭州华为数字技术有限公司 | Physical node upgrading method and device |
-
2017
- 2017-03-30 CN CN201710200566.2A patent/CN108667641A/en active Pending
-
2018
- 2018-03-30 WO PCT/CN2018/081357 patent/WO2018177417A1/en active Application Filing
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140188955A1 (en) * | 2001-06-05 | 2014-07-03 | Silicon Graphics International Corp. | Clustered filesystem with membership version support |
| CN101217678A (en) * | 2007-12-28 | 2008-07-09 | 华为技术有限公司 | System upgrade method, system and device |
| CN101741894A (en) * | 2008-11-26 | 2010-06-16 | ***通信集团公司 | Upgrade method for distributed system and upgrade scheduling node and system |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110247791A (en) * | 2019-05-22 | 2019-09-17 | 北京英视睿达科技有限公司 | The environmental monitoring system and update method of work online updating free of discontinuities |
| CN110247791B (en) * | 2019-05-22 | 2022-04-01 | 北京英视睿达科技股份有限公司 | Environment monitoring system and updating method for continuous work online updating |
| CN110597531A (en) * | 2019-08-05 | 2019-12-20 | 平安科技（深圳）有限公司 | Distributed module upgrading method and device and storage medium |
| CN110597531B (en) * | 2019-08-05 | 2022-11-08 | 平安科技（深圳）有限公司 | Distributed module upgrading method and device and storage medium |
| CN111111161A (en) * | 2019-12-19 | 2020-05-08 | 上海米哈游天命科技有限公司 | Game server repairing method, device, server and medium |
| CN111111161B (en) * | 2019-12-19 | 2023-08-29 | 上海米哈游天命科技有限公司 | Game server repairing method, device, server and medium |
| CN114201549A (en) * | 2020-09-17 | 2022-03-18 | 金篆信科有限责任公司 | Switching method, system, server and storage medium |
| CN114201549B (en) * | 2020-09-17 | 2022-11-25 | 金篆信科有限责任公司 | Switching method, system, server and storage medium |
| CN111930448A (en) * | 2020-09-23 | 2020-11-13 | 南京梦饷网络科技有限公司 | Method, electronic device, and storage medium for service distribution |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2018177417A1 (en) | 2018-10-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108667641A (en) | Distributed system upgrade method, distributed system and updating and management equipment | |
| CN103841155B (en) | A kind of method for downloading software and software download device | |
| WO2014083295A2 (en) | Restoration of network access | |
| EP1931111B1 (en) | Wireless communication apparatus and wireless communication method | |
| EP2114021A1 (en) | Base station control | |
| CN101883355B (en) | Collocation method and system of terminal parameter and terminal management device | |
| CN104750515A (en) | Firmware version upgrading method and system | |
| WO2005083575A1 (en) | Data communication method, data communication system, and communication terminal device | |
| CN104639615A (en) | Data processing method and data processing device in controller cluster | |
| CN108462984A (en) | A kind of method and terminal carrying out access network management | |
| WO2016177127A1 (en) | Single board upgrading method and device | |
| CN108874410B (en) | Patch management method and device | |
| CN106470114A (en) | A kind of collocation method of OTA AKU and device | |
| CN101212340B (en) | Method for restarting control nodes in automatic switching optical network | |
| CN102811470A (en) | Method and device for selecting gateways | |
| CN108028773B (en) | Method and mobility management entity pool for enhancing telecommunication network performance | |
| JP3738378B2 (en) | System configuration setting method and transmission apparatus | |
| CN103596041A (en) | Information prompt method and device | |
| CN108063980A (en) | A kind of TV application software management method, system and device | |
| CN103188087A (en) | Upgrade processing method and upgrade processing device | |
| CN108174417B (en) | Main/standby switching method and device, related electronic equipment and readable storage medium | |
| CN104869582A (en) | Wireless network backup method and system | |
| CN101309209B (en) | Single board software restoring method and single board thereof | |
| CN103596060A (en) | Information prompt method and device | |
| CN112867026B (en) | Networking processing method and 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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20181016 |