CN1322705C - A method of datum plane reset for forwarding equipment - Google Patents
A method of datum plane reset for forwarding equipment Download PDFInfo
- Publication number
- CN1322705C CN1322705C CNB2003101217737A CN200310121773A CN1322705C CN 1322705 C CN1322705 C CN 1322705C CN B2003101217737 A CNB2003101217737 A CN B2003101217737A CN 200310121773 A CN200310121773 A CN 200310121773A CN 1322705 C CN1322705 C CN 1322705C
- Authority
- CN
- China
- Prior art keywords
- plane
- data
- control plane
- reset
- datum
- 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.)
- Expired - Fee Related
Links
Images
Abstract
The present invention discloses a method for resetting a device data forwarding plane, which comprises the following steps: 1, enabling the service treatment part of a control plane incapable of sensing the reset of the data plane; 2, enabling the control plane not to make operations except for a reset operation; 3, saving data stored by the data plane before being reset; 4, resetting the data plane (preferably comprises the replacement of microcodes in a network processor); 5, recovering the data. The method of the present invention has the advantages of short break period of message forwarding, maintenance of dynamic information studied by a system, etc.
Description
Technical field
The present invention relates to the system reset method of the forwarding unit of a kind of router and/or switch, relate in particular to a kind of data plane reset method of forwarding unit.
Technical background
In recent years, the mode of the forwarding unit of part router and switch from exchanging based on core bus originally was to shifting based on the switching network structure.As shown in Figure 1, these equipment are made up of control plane and datum plane usually, and control plane is responsible for controlling the processing and the forwarding of message and low volume data message, and datum plane is responsible for transmitting data message, are used to guarantee the normal discharge of forwarding at a high speed.
The core devices of control plane only has CPU (CPU), and the core devices of datum plane has network processing unit, switching network, network coprocessor, framer etc.The quantity of core devices and complicated processing thereof, the probability that datum plane is gone wrong will be higher than control plane far away.
In frame forwarding unit (comprising frame router and frame-type switch), control plane is arranged in different veneers with datum plane, can handle the datum plane that goes wrong by to the Redundancy Design on data plane and the mode of masterslave switchover.In the boxlike forwarding unit, control plane and datum plane are positioned on the same veneer, generally can only solve datum plane by whole system is resetted and go wrong.The business of whole system operation has just been interrupted like this.Generally speaking, adopt the method that whole system is resetted to solve the problem that datum plane occurs, following shortcoming arranged:
1, it is long that message is transmitted break period, general clock in a measure, even do not wait in tens minutes, viewing system is restarted the configuration restore required time and is decided;
2, original systematic learning to multidate informations such as route, network topology lose, recovering needs the regular hour;
3, at forwarding unit during as access server, original online user's information dropout is restarted the back user and must be authenticated again that reach the standard grade could normal accesses network;
When 4, there is BUG in the microcode in network processing unit, can't online replacement under the situation of non-interrupting service.
Summary of the invention
An object of the present invention is to provide a kind of method for cc equipment data plane reset, use this method under the unbroken situation of business, to reset the data plane.
Another object of the present invention provides a kind of method for cc equipment data plane reset, under the unbroken situation of business, and the microcode in the network processing unit on online replacement data plane.
A further object of the invention provides a kind of method for cc equipment data plane reset, and the detecting data plane is unusual apace to use this method.
For achieving the above object, the invention provides a kind of method for cc equipment data plane reset, may further comprise the steps: (1) makes the step that the Business Processing part of control plane can't the perception data plane resets; (2) make control plane not carry out the reset operation step of operation in addition to the data plane; (3) preserve the step of the data that datum plane preserved before resetting; (4) step that the data plane is resetted; (5) carry out the step that data are recovered.
Preferably, method of the present invention can't further comprise the step that the detecting data plane is unusual before the step of perception data plane resets in the above-mentioned Business Processing part of control plane that makes; The unusual step in described detecting data plane preferably adopts " the handshake message mechanism " between control plane and the datum plane to realize.
Described " handshake message mechanism " is meant that control plane sent the handshake message of appointing in advance every 0.5 second to datum plane, and the microcode of network processing unit is by traffic management device back response message, to keep heartbeat.When continuous 3 handshake message all do not obtain response, think that promptly datum plane is unusual.
Preferably, the step that described in the method for the present invention the data plane is resetted further comprises the step of microcode of the network processing unit on replacement data plane.This step preferably indicates the mode of the array pointer that loads microcode to realize by changing.
Preferably, the described Business Processing part that makes control plane can't the perception data plane resets step and the described step that makes control plane not carry out operation beyond the reset operation to management plane, the CPU by stopping control plane realizing the response of all interruptions.
The preceding data that reset of preserving in the memory of described datum plane are not removed when resetting, and data were kept in the internal memory of control plane before other of preservation resetted.
In the described step that the data plane is resetted, in the step that data plane framer is resetted, comprise the step that the pile function that enables control plane inquiry port status is set.
When technical solution of the present invention went wrong at datum plane, only the reseting data plane did not need the whole system that resets, and therefore following beneficial effect is arranged:
1. it is short that message is transmitted break period, and within 2 seconds, equipment can recover normal forwarding.
2. the multidate information of learning when system moved before resetting is not lost.
3. at box-shaped device during as access server, originally online user's information is not lost.The user does not need to authenticate again and reaches the standard grade, and can not perceive unusual.
Description of drawings
Fig. 1 is based on the composition schematic diagram of the forwarding unit of switching network;
Fig. 2 is the flow chart of the data plane reset step of one embodiment of the present of invention.
Embodiment
Below in conjunction with accompanying drawing, principle of the present invention and preferred embodiment are described in detail.
When datum plane went wrong, its reason may be:
1, the defective of key chip;
2, the defective of the microcode of operation in the key chip (network processor chip);
3, unpredictable reason causes chip to be in abnormality.
The data plane reset technology is by the mode on independent reseting data plane, and the state when datum plane is returned to operate as normal can solve the problem of above datum plane, does not need whole system is resetted.In the method on independent reseting data of the present invention plane, accomplish following some:
1, must make the Business Processing part of control plane can't perceive resetting of datum plane.Otherwise the physical layer that in a single day control plane perceives port connect to disconnect, will the delete interface route; If box-shaped device is during as access server, control plane also can be because of this reason, with user offlines all on this port;
2, must make control plane during data plane reset, management plane not carried out other operation outside the reset operation.Otherwise it is unusual that the business module of control plane can think that datum plane occurs, thereby cause controlling the unusual of message forwarding;
In an embodiment of the present invention, guarantee above 2 points by the interruption of before and after data plane reset, closing, opening the CPU of control plane.
3, must can preserve the data that datum plane was preserved before resetting, thereby after resetting, can correctly recover.In an embodiment of the present invention, by hardware self preservation and auxiliary correct preservation and the recovery that guarantees data of control plane.
The present invention also provides a kind of (such as less than 2 seconds) at short notice, detects the unusual method of datum plane, carries out data plane reset again after unusual detecting datum plane.Otherwise,, could recover if the control plane operation failure such as caused that the user surveys unusually of datum plane then can only continue to carry out the triggering of operation of reaching the standard grade by the user.Adopt the mode of handshake message to carry out the unusual quick detecting of datum plane in the embodiments of the invention.
In addition, for solving because the unusual problem of datum plane that the defective of microcode causes in the network processing unit may need to carry out the replacement of microcode.Promptly when data plane reset, load microcode patch (microcode after the modification problem).In an embodiment of the present invention, by array pointer, the microcode that control loaded is different.
Below in conjunction with Fig. 2, embodiments of the invention are specifically described, listed accompanying drawing only is used for explanation, is not limitation of the present invention.
Under following two kinds of situations, begin to carry out data plane reset:
1, detects datum plane when unusual at control plane, carry out resetting of datum plane.In the present invention, adopt " handshake message mechanism " to carry out the unusual detecting of datum plane." handshake message mechanism " is meant that control plane sent the handshake message of appointing in advance every 0.5 second to datum plane, and the microcode of moving in the datum plane network processing unit is by traffic management device back response message, to keep heartbeat.When continuous 3 handshake message all do not obtain response, think that promptly datum plane is unusual, enter the data plane reset flow process.
2, user's trigger data plane resets.When network operation on the equipment goes wrong and the developer navigates to microcode when having defective, when control plane has the characteristic of patch mechanism, manufacturer can issue patch, by the patch mechanism (mechanism of a kind of online modification problematic operation code of user by control plane, do not need interrupting service), the form of the microcode after the modification defective with the control plane patch is loaded in the system, activates, preserves.By the triggering of order line, carry out data plane reset; When control plane does not have patch when mechanism, can pass through alternate manner, as with microcode store in the memory device of control plane, by resetting of user's trigger data plane, thus the replacement microcode.
Storage chip in the datum plane both can be by the microcode initialization, also can be by the control plane initialization.During the system boot initialization, control plane be responsible for the clearing data content of the storage chip in the plane, as the initial condition of storage chip, microcode can not be removed the content in this chip, otherwise when reset operation, content will be removed by microcode in the storage chip.And because these contents very big (having 1M not wait to 10M usually), and access speed is slow, and when data plane reset, the content in the storage chip is not lost (unless being eliminated), does not preserve these information in the control plane, so must not be removed by microcode.
As shown in Figure 2, in the first embodiment of the present invention, control plane begins to carry out data plane reset when detecting datum plane unusual.At first apply for the memory headroom of control plane, be used to preserve the multidate information of datum plane.The information that these multidate informations comprise is that store in the network processing unit, that store in the network coprocessor, dispose in that store in the traffic management device, that dispose in the switching network, the framer.All of closing control plane are interrupted then, the scene of protected data plane resets.Then call driving function (these functions are finished the effect that makes the datum plane operate as normal), preserve on-the-spot.The information that these field datas comprise is that store in the network processing unit, that store in the network coprocessor, dispose in that store in the traffic management device, that dispose in the switching network, the framer.The information of preserving in the storage chip self is preserved, and the information of preserving in other device is removed by chip self when initialization.
Next step, the network processing unit on reseting data plane, network coprocessor, traffic management device, switching network, framer.
At first, initialization network processing unit.Load networks processor microcode.
Initialization network coprocessor, traffic management device then; According to the configuration information of being preserved, initialization switching network, framer.
Then, recover on-the-spot.The information that these scenes comprise is that store in the network processing unit, that store in the network coprocessor, dispose in that store in the traffic management device, that dispose in the switching network, the framer.
Because after framer (as the PHY device of Ethernet) reinitialized, configuration restore generally needed the time about 1 second, within during this period of time, port is in the connection interrupt status.Can not allow the control plane perception, to avoid the control plane control operation of being correlated with, as the online user of interruptive port being rolled off the production line etc., therefore must enable the pile function of control plane inquiry port status, when this function did not enable, what control plane was inquired about was the time of day of framer, after enabling, the state of the framer of preserving before control plane inquires and resets, and continue to enable about 3 seconds.
At last, open the interruption that control plane is closed, open the scene.
Still as shown in Figure 2, in the second embodiment of the present invention, by resetting of user's trigger data plane.
In the step of initialization network processing unit, judging whether needs to load microcode, and loads the microcode of revising the back patch according to the result who judges.Microcode to be loaded is in the program space of control plane with the form of global variable, and (program space has been preserved the instruction that CPU carries out, the global data of use etc., data segment, code segment or the like are arranged usually) data segment in, when needs load the microcode of revising the back patch, the conversion indication loads the pointer of microcode, loads the microcode of revising the back patch.
Other step is basically the same as those in the first embodiment, so do not give unnecessary details.
It should be noted that, integrated according to chip, the network processing unit of datum plane, framer, switching network, network coprocessor, traffic management device etc. can be independent chip form exist, also can by the network processes chip integrated traffic management, switching network, network association processing capacity, network processes chip, traffic management chip, switching network chip, network association process chip are packaged into one piece of chip on hardware.In addition, if require that higher port density is arranged, the form that the network processes chip can many pieces of chips on hardware exists.For above various forms, data plane reset method of the present invention is all suitable equally.
More than the present invention is described in detail, but those of ordinary skill in the art is to be appreciated that, under situation about not departing from the scope of the present invention with spirit, various improvement, interpolation and replacement all are possible, and all in claim of the present invention institute restricted portion.
Claims (9)
1. method for cc equipment data plane reset may further comprise the steps:
(1) makes the step that the Business Processing part of control plane can't the perception data plane resets;
(2) make control plane not carry out the reset operation step of operation in addition to the data plane;
(3) preserve the step of the data that datum plane preserved before resetting;
(4) step that the data plane is resetted; And
(5) carry out the step that data are recovered.
2. method according to claim 1 is characterized in that, the described Business Processing part of control plane that makes can't further comprise the step that the detecting data plane is unusual before the step of perception data plane resets.
3. method according to claim 2 is characterized in that, the unusual step in described detecting data plane realizes by " handshake message mechanism " between control plane and the datum plane.
4. method according to claim 3, it is characterized in that, described " handshake message mechanism " is meant the handshake message that control plane was appointed to the datum plane transmission in advance every 0.5 second, microcode can be passed through traffic management device back response message, to keep heartbeat, when continuous 3 handshake message all do not obtain response, think that promptly datum plane is unusual.
5. according to each described method of claim 1 to 4, it is characterized in that the described step that the data plane is resetted further comprises the step of microcode of the network processing unit on replacement data plane.
6. method according to claim 5 is characterized in that, the step of the microcode of the network processing unit on described replacement data plane indicates the mode of the array pointer that loads microcode to realize by changing.
7. according to each described method of claim 1 to 4, it is characterized in that, the described Business Processing part that makes control plane can't the perception data plane resets step and the described step that makes control plane not carry out operation beyond the reset operation to management plane, the CPU by stopping described control plane realizes the response of all interruptions.
8. according to each described method of claim 1 to 4, it is characterized in that, the preceding data that reset of preserving in the memory of described datum plane are not removed when resetting, and other data that described datum plane is preserved before resetting are kept in the internal memory of described control plane.
9. according to each described method of claim 1 to 4, it is characterized in that, in the step that the framer of described datum plane is resetted, comprise the step of the pile function that enables control plane inquiry port status.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2003101217737A CN1322705C (en) | 2003-12-23 | 2003-12-23 | A method of datum plane reset for forwarding equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2003101217737A CN1322705C (en) | 2003-12-23 | 2003-12-23 | A method of datum plane reset for forwarding equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1633076A CN1633076A (en) | 2005-06-29 |
| CN1322705C true CN1322705C (en) | 2007-06-20 |
Family
ID=34844264
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2003101217737A Expired - Fee Related CN1322705C (en) | 2003-12-23 | 2003-12-23 | A method of datum plane reset for forwarding equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1322705C (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104978208B (en) * | 2014-04-14 | 2020-05-12 | 新华三技术有限公司 | Hot restart method and device thereof |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20010055463A (en) * | 1999-12-10 | 2001-07-04 | 서평원 | Hardwired Task Scheduler And Scheduling Method In That Task Scheduler |
| US20030138253A1 (en) * | 2002-01-07 | 2003-07-24 | Sungchang Kim | Dynamic wavelength management method in OBS networks |
| US20030191863A1 (en) * | 2001-07-02 | 2003-10-09 | Globespanvirata Incorporated | Communications system using rings architecture |
| CA2428517A1 (en) * | 2002-05-13 | 2003-11-13 | Tropic Networks Inc. | System and method for distributed resource reservation protocol - traffic engineering (rsvp-te) hitless restart in multi-protocol label switching (mpls) network |
-
2003
- 2003-12-23 CN CNB2003101217737A patent/CN1322705C/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20010055463A (en) * | 1999-12-10 | 2001-07-04 | 서평원 | Hardwired Task Scheduler And Scheduling Method In That Task Scheduler |
| US20030191863A1 (en) * | 2001-07-02 | 2003-10-09 | Globespanvirata Incorporated | Communications system using rings architecture |
| US20030138253A1 (en) * | 2002-01-07 | 2003-07-24 | Sungchang Kim | Dynamic wavelength management method in OBS networks |
| CA2428517A1 (en) * | 2002-05-13 | 2003-11-13 | Tropic Networks Inc. | System and method for distributed resource reservation protocol - traffic engineering (rsvp-te) hitless restart in multi-protocol label switching (mpls) network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1633076A (en) | 2005-06-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11729044B2 (en) | Service resiliency using a recovery controller | |
| US11194679B2 (en) | Method and apparatus for redundancy in active-active cluster system | |
| US7434085B2 (en) | Architecture for high availability using system management mode driven monitoring and communications | |
| CN107517110A (en) | Veneer configuration self-recovery method and device in a kind of distributed system | |
| CN111585835B (en) | Control method and device for out-of-band management system and storage medium | |
| CN107203443A (en) | A kind of method and apparatus of the virtual machine High Availabitity based on KVM virtualization | |
| US5600808A (en) | Processing method by which continuous operation of communication control program is obtained | |
| US11258666B2 (en) | Method, device, and system for implementing MUX machine | |
| CN110109772A (en) | A kind of method for restarting of CPU, communication equipment and readable storage medium storing program for executing | |
| CN113507431B (en) | Message management method, device, equipment and machine-readable storage medium | |
| CN1322705C (en) | A method of datum plane reset for forwarding equipment | |
| US8635500B2 (en) | System and method for powering redundant components | |
| US7302607B2 (en) | Two node virtual shared disk cluster recovery | |
| KR20150104435A (en) | Method of performing transition of operation mode for a routing processor | |
| US20170279667A1 (en) | Providing a redundant connection in response to a modified connection | |
| US11812487B2 (en) | Method, device, extender, and computer medium for automatically restoring connection | |
| US20060023627A1 (en) | Computing system redundancy and fault tolerance | |
| CN112217718A (en) | Service processing method, device, equipment and storage medium | |
| CN117278345B (en) | Energy saving method and device applied to network equipment | |
| CN109213446B (en) | Write cache mode switching method, device and equipment and readable storage medium | |
| CN116566804A (en) | Single-point fault avoidance method for intelligent network card hardware unloading in cloud environment | |
| CN117891563A (en) | Control method and device of virtual machine, storage medium and electronic device | |
| JP2023530772A (en) | Operation status switching method, device, active/standby management system and network system | |
| US20090204773A1 (en) | Method of writing device data in dual controller network storage environment | |
| CN113765748A (en) | Method for processing fault of computing node and computer readable storage medium |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20070620 Termination date: 20171223 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |