CN101621417B - Method and exchanger for managing member machine of colony - Google Patents
Method and exchanger for managing member machine of colony Download PDFInfo
- Publication number
- CN101621417B CN101621417B CN2009101640964A CN200910164096A CN101621417B CN 101621417 B CN101621417 B CN 101621417B CN 2009101640964 A CN2009101640964 A CN 2009101640964A CN 200910164096 A CN200910164096 A CN 200910164096A CN 101621417 B CN101621417 B CN 101621417B
- Authority
- CN
- China
- Prior art keywords
- machine
- switch
- state
- clusters
- message
- 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
Images
Abstract
The invention discloses a method and an exchanger for managing a member machine of a colony. The method comprises the following steps: the exchanger receives an adding request message of a commander and judges if the exchanger is the member machine of other colony or not; if the exchanger is the member machine of the other colony, the exchanger judges if the exchanger is in a first state up or a second state down in the other colony; if the exchanger is in the second state down in the other colony, the exchanger sends a recovery request message to the commander; and if the exchanger doses not receive a response message of the commander, the exchanger is converted into a candidacy machine. The invention ensures that the member machine can automatically trigger corresponding operation to modify configuration information only through a series of monitoring judging processes without logging in the member machine.
Description
Technical field
The present invention relates to the communications field, the method and the switch of member machine in a kind of management cluster.
Background technology
Along with the fast development of ethernet technology, the application of Ethernet switch more and more widely still because Ethernet switch quantity many (especially low side devices) is brought certain degree of difficulty to management work, has therefore proposed the solution of cluster management.In cluster management; After candidate switch adding cluster becomes member switch; No matter it is up or down by controlled state, do not allow to add again other clusters and managed by other command machines, show as member machine after the request of receiving other command machines joins message; No matter the state of self is up or down, all can sends refusal and add message to the other side.This has just drawn following problem:
Fig. 1 be according in the cluster management of correlation technique to the processing sketch map of down state member machine.
When the communication of member machine and command machine is broken down,, will self protocol status be shown as down having experienced still not recovery of effective retention time (holdtime).The member machine of this moment can be constantly sends recovery request message to command machine and go in the hope of being added in the cluster again, and communication is in case recover, and the receive orders response message of machine of member machine just can add this cluster recovery up state again; If communication between the two can not recover all the time; Recovery request message can constantly be sent with certain blanking time; If having this moment the command machine of other clusters to want to add it can go whistle for the member; The member machine that this director loses the command machine management phase will have been lost it and add the meaning of cluster by isolated, has lost administering and maintaining of command machine.
If want to add other clusters in this case; Just need log on the member machine member machine is reverted to candidate's machine (becoming member machine because have only candidate's machine to be added cluster by command machine); It is following to take charge of exchange hands-operation step at me: at first the member machine of this down state is forced to be configured to independent machine (promptly having deleted the configurations such as relevant IP, cluster name, command machine information of cluster), be equivalent to the once operation of removing members on command machine; Be candidate's machine with the independent machine manual configuration and then.
There is following weak point in above-mentioned operating procedure: at first; These configuration orders all will log on the member machine and could operate; And in huge network, to find this switch neither be easily, if the command machine in the cluster is paralysed, all member machines that its is hung down all will lose management; Need one one the modification of going to dispose, that will be a great engineering; Secondly; Revising the role of switch in cluster through artificial configuration also is imappropriate way; Because how this disposes all is to come with actual application scenarios according to the variation of network topology; If the member machine state is not down, or do not have the request of other command machines to join message, we also needn't go to change the role of member machine.
Can not be managed by command machine in the cluster of member machine under it to some down state in the correlation technique; Do not allow again to be added by other clusters; Make switch by isolated, can be proposed effective solution at present as yet by the problem of cluster command machine institute management maintenance.
Summary of the invention
Can not be managed by command machine in the cluster of member machine under it to some down state; Do not allow again to be added by other clusters; Make switch by isolated, can not be proposed the present invention by the problem of cluster command machine institute management maintenance, for this reason; Main purpose of the present invention is to provide the method and the switch of member machine in a kind of management cluster, to address the above problem.
To achieve these goals, according to an aspect of the present invention, the method for member machine in a kind of management cluster is provided.
Method according to member machine in the management cluster of the present invention comprises: the request of switch reception command machine joins message and judges whether switch itself is the member machine in other clusters; If the member machine in other clusters, switch judges that itself state in other clusters is the first state up or the second state down; If the state of switch in other clusters is down then sends recovery request message to command machine; The response message of machine if do not receive orders, switch is with itself changing into candidate's machine.
Preferably, after switch judged whether itself is the member machine in other clusters, this method also comprised: the member machine in other clusters, then the switch responds request joins message.
Preferably, after switch judged that itself state in other clusters is the first state up or the second state down, this method also comprised: if the state of switch in other clusters is up then returns refusal to command machine and join message.
Preferably, if the response message of the machine that do not receive orders, switch comprises itself changing into candidate's machine: if the response message of the machine that in the preset time, do not receive orders, switch is with itself changing into candidate's machine.
Preferably, after command machine sent recovery request message, this method also comprised at switch: if the response message of the machine that in the preset time, receives orders, switch adds other cluster recoveries up state again.
Preferably, switch comprises itself changing into candidate's machine: switch transforms order with itself changing into candidate's machine through triggering the role.
Preferably, after switch changed into candidate's machine with itself, this method also comprised: switch sends message to command machine and comes the request of retransferring of notification machine to join message.
To achieve these goals, according to a further aspect in the invention, a kind of switch is provided.
This switch comprises: first judge module is used for judging whether switch itself is the member machine of other clusters; Second judge module is used for judging that this state in other clusters of switch is state one up or state two down; Sending module is used for sending recovery request message to command machine; Conversion module is used for itself changing into candidate's machine.
Preferably, this switch also comprises: trigger module is used for judging that switch originally triggers timer after the state of other clusters is state two down.
Preferably, this switch also comprises: receiver module is used to receive the response message that command machine returns.The 3rd judge module is used to judge whether in the preset time, receive the response message that command machine returns.
Through the present invention, the request of adopting switch to receive command machine joins message and judges whether switch itself is the member machine in other clusters; If the member machine in other clusters, switch judges that itself state in other clusters is the first state up or the second state down; If the state of switch in other clusters is down then sends recovery request message to command machine; The response message of machine if do not receive orders; Switch is with itself changing into candidate's machine; Solved in the cluster of member machine under it of some down state and can not have been managed, do not allowed again to be added, made switch by isolated by other clusters by command machine; Can not be by the problem of cluster command machine institute management maintenance, and then reached and only need judge flow process and need not log on member machine through series of monitoring and just can let member machine self trigger corresponding operation revise the effect of configuration information.
Description of drawings
Accompanying drawing described herein is used to provide further understanding of the present invention, constitutes the application's a part, and illustrative examples of the present invention and explanation thereof are used to explain the present invention, do not constitute improper qualification of the present invention.In the accompanying drawings:
Fig. 1 be according in the cluster management of correlation technique to the processing sketch map of down state member machine;
Fig. 2 is the flow chart according to the method for member machine in the management cluster of the embodiment of the invention;
Fig. 3 is the preferred flow charts that is converted into candidate's machine according to the member machine that makes Down state in the cluster of the embodiment of the invention;
Fig. 4 is the sketch map according to the preferred switch of the embodiment of the invention.
Embodiment
Function was both stated
Consider in the cluster of member machine under it of some down state and can not be managed by command machine; Do not allow again to be added by other clusters; Make switch by isolated; Can not be by cluster command machine institute management maintenance, the embodiment of the invention provides the method and the switch of member machine in a kind of management cluster, and this method comprises: the request of switch reception command machine joins message and judges whether switch itself is the member machine in other clusters; If the member machine in other clusters, switch judges that itself state in other clusters is the first state up or the second state down; If the state of switch in other clusters is down then sends recovery request message to command machine; The response message of machine if do not receive orders, switch is with itself changing into candidate's machine.
Preferably; Equipment is after the request of the machine of receiving orders joins message; Note medium access control (the Media Access Control of command machine earlier; Abbreviate MAC as) address; Judge again through MAC Address whether oneself has been the member machine in other cluster, if words judge that again the state of this member machine in cluster is up or down, join message and get final product if the words of up are just directly returned refusal; If the words of down just trigger a timer (this timer can be set to 10 minutes: prevent command machine because that situation such as power-off restarting causes is unreachable); In timer time, continue to send recovery request message to former command machine, if behind timer expiry, do not receive the response message (being that the far-end command machine is unreachable) of former command machine yet, member machine will trigger changing role with oneself changing into candidate's machine; About the relevant configuration of cluster, send the flush message to the command machine that previous transmission request joins message then: notify its repeat requests to join message before removing.For preventing the flush message dropping, can be provided with and send number of times is 3, is spaced apart 8s (with reference to the transmission blanking time of command machine in the cluster and member machine handshake message).Command machine retransmits previous request again and joins message after receiving such flush message, be exactly the normal adition process of cluster next.So just can under the situation of not landing member machine modification configuration, manage again equipment.
If but the down member machine do not receive that other command machines join message and just do not need any operation, continue to send recovery request message till communication recovers.
Need to prove that under the situation of not conflicting, embodiment and the characteristic among the embodiment among the application can make up each other.Below with reference to accompanying drawing and combine embodiment to specify the present invention.
Method embodiment
According to embodiments of the invention, the method and the switch of member machine in a kind of management cluster is provided.
Fig. 2 is the flow chart according to the method for member machine in the management cluster of the embodiment of the invention.
As shown in Figure 2, this method comprises that following step S202 is to step S208:
Step S202, the request of switch reception command machine joins message and judges whether switch itself is the member machine in other clusters;
Step S204, if the member machine in other clusters, switch judges that itself state in other clusters is the first state up or the second state down;
Step S206 is if the state of switch in other clusters is down then sends recovery request message to command machine;
Step S208, if the response message of the machine that do not receive orders, switch is with itself changing into candidate's machine.
To combine instance that the implementation procedure of the embodiment of the invention is described in detail below.
Fig. 3 is the preferred flow charts that is converted into candidate's machine according to the member machine that makes Down state in the cluster of the embodiment of the invention.
As shown in Figure 3, it is following that the present invention makes the member machine of down state in the cluster be converted into the treatment step of candidate's machine technology scheme flow process part:
Step 1: the request that equipment receives command machine joins message.
Step 2: whether judgment device has been the member machine in other clusters, judges that the state of this member machine in cluster is up or down if change step 3 over to, otherwise changes step 5 over to, gets into normal cluster management resume module flow process.
Step 3: judge that the state of this member machine in cluster is up or down,, get into normal cluster management resume module flow process if up changes step 5 over to; If down changes step 4 over to.
Step 4: just trigger a timer if member machine is down; In this timer time, continue to send recovery request message to former command machine; If before timer expiry, just can receive the response message of former command machine; Then change step 5 over to, get into normal cluster management resume module flow process; Otherwise change step 6 over to.
Step 5: get into normal cluster management resume module flow process; If being up, member machine joins message with regard to directly returning refusal; Transfer to the cluster management resume module; If before timer expiry, just can receive the response message of former command machine, member machine just can add this cluster recovery up state again, gets into the handling process of up state.
Step 6: show that former command machine is unreachable, trigger the role and transform order,, send its request before that retransmits of flush message notifying to new command machine and join message with self being converted into candidate's machine.Preferably; If behind timer expiry, do not receive the response message of former command machine yet; Member machine self will trigger the role and transform order; Get into the handling process under the down pattern: at first the removing members operation as the machine of receiving orders becomes candidate's machine with self-recovery, and the command machine that joins message to previous transmission request then sends the previous request of its re-transmission of flush message notifying and joins message.
From above description; Can find out that the present invention has realized following technique effect: adopt the present invention to make the member machine of down state in the cluster be converted into the method for candidate's machine, revise configuration information on the member machine and compare with existing landing: one of which; Saved the processing time; Improved operating efficiency, need not log on member machine place frame and realize that through revising configuration the present invention only need judge that flow process just can let member machine self trigger corresponding operation achieve the goal through series of monitoring.Its two, can be able to handle flexibly to different environment, have only the member machine ability needs trigger command that has satisfied some condition to make it be converted into candidate's machine, the role that other any situation all can not the hack switch.
Device embodiment
According to embodiments of the invention, a kind of switch is provided.
Fig. 4 is the sketch map according to the preferred switch of the embodiment of the invention.
As shown in Figure 4, this switch 400 comprises: first judge module 402, second judge module 404, sending module 406 and conversion module 408.
Particularly, first judge module 402 is used for judging whether switch itself is the member machine of other clusters; Second judge module 404 is used for judging that this state in other clusters of switch is state one up or state two down; Sending module 406 is used for sending recovery request message to command machine; Conversion module 408 is used for itself changing into candidate's machine.
Preferably, switch 400 also comprises: trigger module 410 is used for judging that switch originally triggers timer after the state of other clusters is state two down.
Preferably, switch 400 also comprises: receiver module 412 is used to receive the response message that command machine returns.
Preferably, switch 400 also comprises: the 3rd judge module 414 is used to judge whether in the preset time, receive the response message that command machine returns.
More than be merely the preferred embodiments of the present invention, be not limited to the present invention, for a person skilled in the art, the present invention can have various changes and variation.All within spirit of the present invention and principle, any modification of being done, be equal to replacement, improvement etc., all should be included within protection scope of the present invention.
Claims (10)
1. the method for member machine in the management cluster is characterized in that, comprising:
The request of switch reception command machine joins message and judges whether said switch itself is the member machine in other clusters;
If the member machine in other clusters, said switch judges that itself state in said other clusters is the first state up or the second state down;
If the state of said switch in said other clusters is down then sends recovery request message to former command machine;
If do not receive the response message of said former command machine, said switch is with itself changing into candidate's machine.
2. method according to claim 1 is characterized in that, after said switch judged whether itself is the member machine in other clusters, said method also comprised:
Member machine in other clusters, then said switch responds described request joins message.
3. method according to claim 1 is characterized in that, after said switch judged that itself state in said other clusters is the first state up or the second state down, said method also comprised:
If the state of said switch in said other clusters be up to said command machine return the refusal join message.
4. method according to claim 1 is characterized in that, if do not receive the response message of said former command machine, said switch comprises itself changing into candidate's machine:
If in the preset time, do not receive the response message of said former command machine, said switch is with itself changing into candidate's machine.
5. method according to claim 4 is characterized in that, after said former command machine sent recovery request message, said method also comprised at said switch:
If in the preset time, receive the response message of said former command machine, said switch adds said other cluster recoveries up state again.
6. method according to claim 4 is characterized in that, said switch comprises itself changing into candidate's machine:
Said switch transforms order with itself changing into candidate's machine through triggering the role.
7. method according to claim 1 is characterized in that, after said switch changed into candidate's machine with itself, said method also comprised:
Said switch sends message to said command machine and notifies the request of retransferring of said command machine to join message.
8. a switch is characterized in that, comprising:
First judge module, the request that is used for receiving command machine joins message and judges whether said switch itself is the member machine of other clusters;
Second judge module is used for judging that this state in said other clusters of said switch is state one up or state two down;
Sending module is used at said switch when the state of said other clusters is down, sends recovery request message to former command machine;
Conversion module is used for when not receiving the response message of said former command machine, with itself changing into candidate's machine.
9. switch according to claim 8 is characterized in that, said switch also comprises:
Trigger module is used for judging that said switch originally triggers timer after the state of said other clusters is state two down.
10. switch according to claim 9 is characterized in that, said switch also comprises:
Receiver module is used to receive the response message that said former command machine returns;
The 3rd judge module is used to judge whether in the preset time, receive the response message that said former command machine returns.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009101640964A CN101621417B (en) | 2009-08-11 | 2009-08-11 | Method and exchanger for managing member machine of colony |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2009101640964A CN101621417B (en) | 2009-08-11 | 2009-08-11 | Method and exchanger for managing member machine of colony |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101621417A CN101621417A (en) | 2010-01-06 |
| CN101621417B true CN101621417B (en) | 2012-01-11 |
Family
ID=41514476
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2009101640964A Active CN101621417B (en) | 2009-08-11 | 2009-08-11 | Method and exchanger for managing member machine of colony |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101621417B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6636499B1 (en) * | 1999-12-02 | 2003-10-21 | Cisco Technology, Inc. | Apparatus and method for cluster network device discovery |
| CN1484412A (en) * | 2002-09-20 | 2004-03-24 | 华为技术有限公司 | Method for realizing 802.1 X communication based on group management |
| CN1525690A (en) * | 2003-09-18 | 2004-09-01 | 港湾网络有限公司 | A method for implementing remote cluster management on low end switch |
| CN101238684A (en) * | 2005-09-12 | 2008-08-06 | 中兴通讯股份有限公司 | System for cluster managing in the Ethernet switch layer and the method thereof |
-
2009
- 2009-08-11 CN CN2009101640964A patent/CN101621417B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6636499B1 (en) * | 1999-12-02 | 2003-10-21 | Cisco Technology, Inc. | Apparatus and method for cluster network device discovery |
| CN1484412A (en) * | 2002-09-20 | 2004-03-24 | 华为技术有限公司 | Method for realizing 802.1 X communication based on group management |
| CN1525690A (en) * | 2003-09-18 | 2004-09-01 | 港湾网络有限公司 | A method for implementing remote cluster management on low end switch |
| CN101238684A (en) * | 2005-09-12 | 2008-08-06 | 中兴通讯股份有限公司 | System for cluster managing in the Ethernet switch layer and the method thereof |
Non-Patent Citations (1)
| Title |
|---|
| 黄旭熙，游松.以太网交换机集群与组网.《计算机工程与设计》.2005,第26卷(第12期),3288-3290. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101621417A (en) | 2010-01-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100466584C (en) | Processing method for loop network protection | |
| CN105227385B (en) | A kind of method and system of troubleshooting | |
| JP2022502926A (en) | UE migration method, equipment, system, and storage medium | |
| CN101060485B (en) | Topology changed messages processing method and processing device | |
| US20140254347A1 (en) | Ethernet Ring Protection Switching Method, Node, and System | |
| CN104935672A (en) | High available realizing method and equipment of load balancing service | |
| CN103200109B (en) | A kind of ospf neighbor relationship management method and equipment | |
| CN109218126B (en) | Method, device and system for monitoring node survival state | |
| CN102035676A (en) | ARP (Address Resolution Protocol) interaction based method and equipment for detecting and recovering link fault | |
| CN102780593B (en) | Based on the method for BFD protocol detection link, device and network processing unit | |
| CN104471915A (en) | Grace state and pacing in link aggregation | |
| CN101674235B (en) | Data transmission method and apparatus | |
| CN103167479B (en) | Realize that terminal is distant gets killed or the distant methods, devices and systems opened | |
| CN102215378A (en) | Multimedia sensing network system capable of realizing dynamic networking | |
| CN105790985A (en) | Data switching method, first device, second device and device | |
| CN101621417B (en) | Method and exchanger for managing member machine of colony | |
| CN100413248C (en) | Improved method and system for carrying out charging based on flow | |
| CN113794632B (en) | Network protection method and device | |
| WO2012151855A1 (en) | Method and system for managing cloud computing device | |
| CN110505104B (en) | Fault processing method, device, network equipment and storage medium | |
| CN104301240B (en) | Data transmission method and system | |
| CN111224803B (en) | Multi-master detection method in stacking system and stacking system | |
| CN101383718B (en) | Method and device for network management system to share a part of pause rule | |
| CN103037502B (en) | A kind of method and system of processing terminal triggering message | |
| CN102833093A (en) | Network fault processing method, device and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |