CN1949699B - Method and apparatus for reducing single-board fault affect on service - Google Patents
Method and apparatus for reducing single-board fault affect on service Download PDFInfo
- Publication number
- CN1949699B CN1949699B CN2006101383694A CN200610138369A CN1949699B CN 1949699 B CN1949699 B CN 1949699B CN 2006101383694 A CN2006101383694 A CN 2006101383694A CN 200610138369 A CN200610138369 A CN 200610138369A CN 1949699 B CN1949699 B CN 1949699B
- Authority
- CN
- China
- Prior art keywords
- veneer
- mainboard
- traffic lost
- lost rate
- service
- 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
Landscapes
- Monitoring And Testing Of Exchanges (AREA)
Abstract
The invention provides a method and device for reducing the effect of single plate failure on service, belonging to communication technical field. And the method comprises steps of: calculating fail service loss rate, calculating master plate service loss rate and backup plate service loss rate, and switching or not according to the master and backup plate service loss rates. And the device comprises operation module and processing module. And the invention is simple and easy to operate, having high calculating efficiency, reducing system complexity and effectively avoiding ping-pong switching occurrence, and assuring the service ability the system offers is the greatest; has wide application range and extensibility.
Description
Technical field
The present invention relates to communication technique field, particularly a kind of method and apparatus that reduces effect of single plate failure on service.
Background technology
In order to improve the reliability of system, when design realization system, can carry out Redundancy Design for the veneer of key.Present modal veneer Redundancy Design is the main and standby boards backup scenario, and two veneers, activestandby states are each other promptly arranged in the system.When normal operation, a veneer wherein externally provides normal function with professional as mainboard; Another piece veneer is as slave board, and its moves but does not externally provide professional, but the various data and the state of the mainboard of backing up in realtime externally provide professional so that can substitute former mainboard smoothly in needs.
When oneself breaking down (generally being serious or critical failure) when mainboard detects in running, system is thought, and the function and business that externally provides can be provided, then mainboard carries out switching operation automatically, it is mainboard notice slave board, allow slave board upgrade to mainboard and externally provide professional, oneself reduce to slave board simultaneously, as the backup veneer of new mainboard, back up the data and the state of new mainboard, get ready for switching next time.
Such scheme can expand to the N+1 scheme, and promptly N piece veneer all externally provides professional as mainboard, and 1 veneer does not externally provide professional as slave board.In fact, the main and standby boards backup scenario can be regarded the special case of N+1 scheme, i.e. N=1 as.
The shortcoming of above-mentioned prior art is as follows:
1, do not check the state of slave board before mainboard initiates to switch, and slave board is when breaking down also just, the situation that can occur rattles switches, the system that not only has a strong impact on provide professional ability, have also reduced the stability of system;
When all there was fault in main and standby boards, because mainboard finds oneself to exist the fault decision to switch, and former slave board found oneself also to exist fault after upgrading to mainboard, initiates to switch decision-making again.Can cause two veneers ceaselessly carrying out switching operation like this, table tennis generally take place this moment switched.
Even check the state of slave board before 2 mainboards initiate to switch, also just fault is analyzed qualitatively, rather than quantitative analysis, very likely can reduce system can externally provide professional specification;
When all there is fault in active and standby plate, if the degree of slave board fault is compared lighter with the degree of main board failure, for example the fault degree of mainboard is seriously to can't externally providing professional, though slave board also has fault, but can also provide professional, then mainboard only draws the conclusion that also there is fault in slave board according to analyzing qualitatively, and do not switch, will reduce system professional specification externally is provided, even the veneer that occurs providing professional is as mainboard, and slave board can provide the situation of more business.
Also having a kind of common veneer Redundancy Design is the resource pool scheme, and promptly the polylith veneer is formed a resource pool, and the as a whole business that externally provides is provided, and the general the lightest veneer of business load of selecting when carrying out traffic assignments comes bearer service.When certain piece single board default wherein, can withdraw from resource pool, no longer externally provide professional.This scheme does not consider that veneer has the situation of traffic lost, when distribution service, only considers the ability of load and the situation of actual loading, and is accurate inadequately.
Summary of the invention
May to occur table tennis in the prior art and switch and problem such as traffic profile reduction in order to solve, the invention provides a kind of method that reduces effect of single plate failure on service, specifically may further comprise the steps:
The various calculation of fault that occur according to veneer have various fault traffic lost rates;
All fault traffic lost rate summations to described veneer obtain the single board service loss late; Wherein, described veneer comprises mainboard and slave board;
Whether switch described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer.
Described fault traffic lost rate is calculated as follows:
The total professional ability that provides professional ability/veneer to provide of veneer loss is provided fault traffic lost rate=fault.
Described step of whether switching described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer is specially:
When mainboard traffic lost rate in the described veneer greater than described veneer in during slave board traffic lost rate, switch described mainboard and slave board; Otherwise do not switch.
Set in advance a threshold value, described step of whether switching described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer is specially:
When mainboard traffic lost rate in the described veneer greater than described veneer in slave board traffic lost rate, and described mainboard traffic lost rate is switched described mainboard and slave board during also more than or equal to described preset threshold value; Otherwise do not switch.
Set in advance a hysteresis, described step of whether switching described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer is specially:
When the difference of slave board traffic lost rate in mainboard traffic lost rate and the veneer in the described veneer during, switch described mainboard and slave board greater than described default hysteresis; Otherwise do not switch.
Set in advance a threshold value and a hysteresis, described step of whether switching described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer is specially:
The difference of slave board traffic lost rate is greater than described default hysteresis in mainboard traffic lost rate and veneer in the described veneer, and described mainboard traffic lost rate is switched described mainboard and slave board during more than or equal to described preset threshold value; Otherwise do not switch.
The present invention also provides a kind of device that reduces effect of single plate failure on service, and described device comprises:
Computing module is used for haveing various fault traffic lost rates according to the various calculation of fault that veneer occurs; Also be used for all the fault traffic lost rate summations to described veneer, obtain the single board service loss late, wherein, described veneer comprises mainboard and slave board;
Processing module is used for whether switching described mainboard and slave board according to slave board traffic lost rate decision in described veneer mainboard traffic lost rate and the veneer.
Described computing module comprises:
Fault traffic lost rate arithmetic element is used to be calculated as follows out various fault traffic lost rates:
The total professional ability that provides professional ability/veneer to provide of veneer loss is provided fault traffic lost rate=fault;
Single board service loss late arithmetic element is used for all the fault traffic lost rate summations to described veneer, obtains the single board service loss late, and wherein, described veneer comprises mainboard and slave board.
Described processing module comprises:
Whether first processing unit is used for judging described veneer mainboard traffic lost rate greater than slave board traffic lost rate in the described veneer, if greater than, then switch described mainboard and slave board; Otherwise do not switch;
Or second processing unit, be used to set in advance a threshold value, when mainboard traffic lost rate in the described veneer greater than described veneer in slave board traffic lost rate, and described mainboard traffic lost rate is switched described mainboard and slave board during also more than or equal to described preset threshold value; Otherwise do not switch;
Or the 3rd processing unit, be used to set in advance a hysteresis, when the difference of slave board traffic lost rate in mainboard traffic lost rate and the veneer in the described veneer during, switch described mainboard and slave board greater than described default hysteresis; Otherwise do not switch;
Or manage the unit everywhere, be used to set in advance a threshold value and a hysteresis, the difference of slave board traffic lost rate is greater than described default hysteresis in mainboard traffic lost rate and veneer in the described veneer, and described mainboard traffic lost rate is switched described mainboard and slave board during more than or equal to described preset threshold value; Otherwise do not switch.
In order to solve the accurate inadequately problem of veneer load allocating in the prior art, the present invention also provides a kind of method that reduces effect of single plate failure on service, and described method specifically may further comprise the steps:
The various calculation of fault that occur according to veneer have various fault traffic lost rates;
All fault traffic lost rate summations to described veneer obtain the single board service loss late;
Calculate the veneer load factor according to described single board service loss late, judge then whether described veneer load factor is veneer load factor minimum in the system, if then give this veneer distribution service, otherwise do not give this veneer distribution service.
Described veneer load factor is calculated as follows: when described single board service loss late less than 100% the time,
Veneer load factor=veneer current business amount/(single board service total amount * (1-single board service loss late));
When described single board service loss late equaled 100%, getting the veneer load factor was 100%.
The present invention also provides a kind of device that reduces effect of single plate failure on service, and described device comprises:
Computing module is used for haveing various fault traffic lost rates according to the various calculation of fault that veneer occurs; Also be used for all fault traffic lost rate summations, obtain the single board service loss late described veneer; Also be used for calculating the veneer load factor according to described single board service loss late;
Processing module is used to judge whether described veneer load factor is veneer load factor minimum in the system, if then give this veneer distribution service, otherwise do not give this veneer distribution service.
Described computing module comprises:
Veneer load factor arithmetic element is used to be calculated as follows out the veneer load factor:
When described single board service loss late less than 100% the time,
Veneer load factor=veneer current business amount/(single board service total amount * (1-single board service loss late));
When described single board service loss late equaled 100%, getting the veneer load factor was 100%.
Beneficial effect of the present invention mainly shows:
1, by calculating to mainboard traffic lost rate and slave board traffic lost rate, whether carry out switching of active and standby plate with traffic lost rate decision according to active and standby plate, rather than whether break down according to veneer simply and determine whether carrying out switching of active and standby plate, the situation of not only having avoided table tennis to switch effectively takes place, but also guaranteed that the professional ability that system externally provides is maximum, avoided the low veneer of professional ability also externally to provide professional situation to take place effectively as mainboard.
2, applicable scope of the present invention is wide, by calculating mainboard traffic lost rate and slave board traffic lost rate, and judges whether to switch according to corresponding condition, and the present invention both can be applied on the main and standby boards backup scenario, also can be applied on the N+1 scheme; By calculating, and carry out traffic assignments, the present invention can also be applied on the resource pool scheme according to corresponding condition to single board service loss late and veneer load factor.
Beneficial effect of the present invention further comprises:
3, technical solution of the present invention is simple, understands easily, and computing formula is uncomplicated, the computational efficiency height, and wherein the single board service loss late can reflect the state of current single board exactly.
4, mainboard and slave board only need the information of switching board traffic lost rate among the present invention, and do not need the situation of the concrete fault of switching board, and the amount of information of required exchange is few, has reduced the complexity of system.
5, the present invention has extensibility, and different product or different veneer can be according to the situations of oneself, the traffic lost rate computing formula of expanding single plate and judgement formula on basis of the present invention.
Description of drawings
Fig. 1 is the flow chart of the embodiment one of the present invention's method of reducing effect of single plate failure on service;
Fig. 2 is the structure chart of the embodiment two of the present invention's device of reducing effect of single plate failure on service;
Fig. 3 is the flow chart of the embodiment three of the present invention's method of reducing effect of single plate failure on service;
Fig. 4 is the structure chart of the embodiment four of the present invention's device of reducing effect of single plate failure on service.
Embodiment
The invention will be further described below in conjunction with the drawings and specific embodiments, but not as a limitation of the invention.
The present invention considers the situation of veneer partial fault, has proposed the notion of fault traffic lost rate, when the veneer complete failure, can think the special case of veneer partial fault, and promptly the traffic lost rate is 100%.By quantitative analysis, provide to mainboard and to switch decision-making accurately and effectively, to improve the reliability of board switchover, guarantee that as much as possible veneer externally provides professional ability, reduces the influence that single board default causes business.For the resource pool scheme, calculating by fault traffic lost rate, single board service loss late and veneer load factor, can calculate the veneer of system's internal burden minimum more accurately, and give the veneer distribution service of veneer load factor minimum, thereby reduce the influence of single board default business.
Embodiment one
System in the present embodiment adopts the main and standby boards backup scenario, two veneers is promptly arranged, a mainboard and a slave board in the system.Referring to Fig. 1, the invention provides a kind of method that reduces effect of single plate failure on service, specifically may further comprise the steps:
Step 101: the various faults that occur according to veneer provide professional influence to veneer, calculate the fault traffic lost rate of every kind of fault;
The computing formula of fault traffic lost rate LF is:
The total professional ability that provides professional ability/veneer to provide of veneer loss is provided fault traffic lost rate LF=fault.
The professional ability that veneer provides is in different application scenarioss, and the unit of calculating is also different.For example when veneer provides voice service, calculate the professional ability of veneer with number of users; When veneer provides the service of transmit IP message, calculate the professional ability of veneer with the bandwidth that provides, as 100MB.
The traffic lost rate of different faults realizes and difference according to the design of dissimilar veneers, promptly needs to determine according to the dissimilar veneer of the different product of reality.For example, the veneer of supposing certain type externally provides the professional ability at 4 general-purpose families, externally provides professional ability jointly by 4 DSP, if wherein a DSP breaks down, then the professional ability that provides of this veneer loss is 1 power user, so the fault traffic lost rate of this fault is 1/4, promptly 25%.
In the present embodiment, the veneer in the supposing the system exists four kinds business had the fault of influence in various degree, and is as shown in table 1:
| The fault title | Fault traffic lost rate LF (reducing the percentage of traffic profile) |
| F1 | 100％ |
| F2 | 50％ |
| F3 | 20％ |
| F4 | 10％ |
Table 1
When the F1 fault appearred in veneer, veneer can lose 100% professional ability, and promptly veneer can't externally provide professional; When the F4 fault appearred in veneer, then veneer can lose 10% professional ability, and promptly veneer can also provide 90% professional ability.Make the many more faults of veneer lost traffic, big more to the veneer influence, illustrate that this fault is serious more.
The calculating of fault traffic lost rate can be calculated when needing at every turn again, in order to improve the efficient of calculating, also can the accumulative total formula calculate when having new fault to produce or recovering.It is exactly on the basis of former fault traffic lost rate that the accumulative total formula is calculated, the fault traffic lost rate that the new fault that produces that adds up causes, and the fault traffic lost rate that causes of the fault that deducts new recovery, as current new fault traffic lost rate, concrete formula is with the result:
The fault traffic lost rate that the fault that the fault traffic lost rate that new fault traffic lost rate=former fault traffic lost rate+new fault causes-Xin recovers causes.
Step 102: the various fault traffic lost rate summation operation to mainboard, obtain mainboard traffic lost rate, the various fault traffic lost rate summation operation to slave board obtain slave board traffic lost rate; Promptly be calculated as follows:
Single board service loss late LB=∑ LFi (1＜i＜M);
This formula all is suitable for for mainboard and slave board; Wherein M is the fault sum that occurs on this veneer, and the fault traffic lost rate of i fault correspondence is LFi, the addition of M on the veneer fault traffic lost rate, as the total traffic lost rate LB of this veneer.For example, as if the fault F3 and the F4 that occur simultaneously on the veneer in the table 1, then single board service loss late LB=LF3+LF4=20%+10%=30%.
When the total loss of veneer surpasses 100%, can think that then veneer externally can not provide professional, so among the present invention the traffic lost rate that surpasses 100% all be thought 100%.For example, if fault F1 and F2 occur simultaneously, then the single board service loss late is 100%+50%=150%; If fault F1 and F4 occur simultaneously, then the single board service loss late is 100%+10%=110%; Though the result who calculates is inequality, it is considered herein that they are identical to the influence of veneer, so be 100% all with single board service loss late value.According to this situation, above-mentioned formula is revised as:
Single board service loss late LB=min (∑ LFi, 100%).
Because veneer comprises mainboard and slave board, so the plate traffic lost rate of calculating in this step just comprises mainboard traffic lost rate and slave board traffic lost rate.
Step 103: whether judge mainboard traffic lost rate greater than slave board traffic lost rate, if then in system, carry out the switching operation of mainboard and slave board; Otherwise do not switch; When promptly satisfying following condition, just switch:
(mainboard traffic lost rate LB-slave board traffic lost rate LB)＞0;
For example, if slave board traffic lost rate is 50%, and mainboard traffic lost rate is 20%, then since mainboard traffic lost rate less than slave board traffic lost rate, so can not carry out masterslave switchover.If slave board traffic lost rate is 50%, and mainboard traffic lost rate is 100%, and masterslave switchover then can take place, switches the new mainboard in back 50% professional ability externally is provided.
When mainboard traffic lost rate equates with slave board traffic lost rate, do not switch, can avoid table tennis to switch like this.
Consider the service disconnection that board switchover might cause some providing, and the veneer most of the time all expires the specification operation, so sometimes user's specification of would rather accepting business reduces, and the service disconnection of being unwilling.Therefore can set in advance a threshold value, whether the judgment condition of board switchover further increases judges mainboard traffic lost rate LB more than or equal to described threshold value, as follows:
(mainboard traffic lost rate LB 〉=threshold value) AND (mainboard traffic lost rate LB-slave board traffic lost rate LB＞0);
Promptly have only when mainboard traffic lost rate more than or equal to certain pre-set threshold value, and mainboard traffic lost rate is just carried out the switching operation of mainboard and slave board during greater than slave board traffic lost rate.The purpose that threshold value is set is exactly to reduce the service disconnection that board switchover causes, has only mainboard traffic lost rate to surpass or just can switch when equaling threshold value.
Above-mentioned threshold value can be static in advance the setting, also can dynamically adjust according to the situation of veneer operation.For example, suppose that preset threshold value is 30%, through learning after the statistics that veneer its traffic carrying capacity in running never surpasses 50%, in this case, even the single board service loss late reaches 40%, it can also provide 60% professional ability so, and actual volume can be greater than 50%, therefore do not switch and to guarantee that veneer normally moves yet, then the threshold value of traffic lost rate can be adjusted into 40% from 30% this moment, like this as long as the traffic lost in the threshold value control scope can not influence the normal operation of system.
Further,, in above-mentioned judgment condition, increase a hysteresis in order to reduce the frequency of board switchover, promptly below satisfied during condition, just allow to switch:
(mainboard traffic lost rate LB-slave board traffic lost rate LB)＞hysteresis;
Then above-mentioned formula is modified to:
(mainboard traffic lost rate LB 〉=threshold value) AND (mainboard traffic lost rate LB-slave board traffic lost rate LB＞hysteresis);
For example, mainboard traffic lost rate is 50%, slave board traffic lost rate is 30%, threshold value is 40%, hysteresis is 10%, and then mainboard traffic lost rate is greater than threshold value 40%, and the difference of mainboard traffic lost rate and slave board traffic lost rate is: 50%-30%=20%, the result switches so need carry out mainboard greater than hysteresis 10%.Do not increase the formula of hysteresis, the formula before promptly revising can be regarded hysteresis as and be 0 o'clock special case.
In sum, present embodiment can adopt in following four kinds of judgment condition any to determine whether initiating active and standby plate to switch:
(1) (mainboard traffic lost rate LB-slave board traffic lost rate LB)＞0;
(2) (mainboard traffic lost rate LB 〉=threshold value) AND (mainboard traffic lost rate LB-slave board traffic lost rate LB＞0);
(3) (mainboard traffic lost rate LB-slave board traffic lost rate LB)＞hysteresis;
(4) (mainboard traffic lost rate LB 〉=threshold value) AND (mainboard traffic lost rate LB-slave board traffic lost rate LB＞hysteresis).
Present embodiment is the main and standby boards backup scenario, when system adopts the N+1 scheme, N mainboard is arranged in the system, a slave board, and then above-mentioned formula can be revised as:
(mainboard traffic lost rate LBj 〉=threshold value) AND (mainboard traffic lost rate LBj-slave board traffic lost rate LB＞0), (0＜j＜N);
Wherein LBj represents the traffic lost rate of j mainboard in the N of the system mainboard, then for certain mainboard, only in its traffic lost rate more than or equal to threshold value, and when bigger, just carry out switching operation than the value of slave board traffic lost rate.In fact, can think that the main and standby boards backup scenario is the special case of N+1 scheme, i.e. N=1.
When a plurality of mainboards of appearance satisfy judgment condition simultaneously, usually get the mainboard of traffic lost rate maximum and switch, for example, have its traffic lost rate of three mainboards to be respectively 100%, 50% and 30%, they all satisfy judgment condition, are that 100% mainboard is switched to the traffic lost rate then.
Embodiment two
Referring to Fig. 2, the invention provides a kind of device that reduces effect of single plate failure on service, specifically comprise:
(1) computing module is used for haveing various fault traffic lost rates according to the various calculation of fault that veneer occurs, and veneer comprises mainboard and slave board; Also be used for the various fault traffic lost rates of mainboard are carried out computing, obtain mainboard traffic lost rate, the various fault traffic lost rates of slave board are carried out computing, obtain slave board traffic lost rate;
(2) processing module is used for whether switching mainboard and slave board according to mainboard traffic lost rate and the decision of slave board traffic lost rate.Computing module comprises:
1) fault traffic lost rate arithmetic element is used to be calculated as follows out various fault traffic lost rates:
The total professional ability that provides professional ability/veneer to provide of veneer loss is provided fault traffic lost rate=fault;
2) single board service loss late arithmetic element is used for all the fault traffic lost rate summations to mainboard, obtains mainboard traffic lost rate; All fault traffic lost rate summations to slave board obtain slave board traffic lost rate.
Processing module comprises:
Whether first processing unit is used to judge mainboard traffic lost rate greater than slave board traffic lost rate, if greater than, then switch mainboard and slave board; Otherwise do not switch.When promptly satisfying following condition, just switch:
(mainboard traffic lost rate-slave board traffic lost rate)＞0.
Or processing module comprises:
Second processing unit is used to set in advance a threshold value, when mainboard traffic lost rate greater than slave board traffic lost rate, and mainboard traffic lost rate is switched mainboard and slave board during also more than or equal to preset threshold value; Otherwise do not switch.When promptly satisfying following condition, just switch:
(mainboard traffic lost rate 〉=threshold value) AND (mainboard traffic lost rate-slave board traffic lost rate＞0).
Or processing module comprises:
The 3rd processing unit is used to set in advance a hysteresis, when the difference of mainboard traffic lost rate and slave board traffic lost rate during greater than default hysteresis, switches mainboard and slave board; Otherwise do not switch.When promptly satisfying following condition, just switch:
(mainboard traffic lost rate-slave board traffic lost rate)＞hysteresis.
Or processing module comprises:
The manages the unit everywhere, is used to set in advance a threshold value and a hysteresis, when the difference of mainboard traffic lost rate and slave board traffic lost rate greater than default hysteresis, and mainboard traffic lost rate is switched mainboard and slave board during more than or equal to preset threshold value; Otherwise do not switch.When promptly satisfying following condition, just switch:
(mainboard traffic lost rate 〉=threshold value) AND (mainboard traffic lost rate-slave board traffic lost rate＞hysteresis).
Embodiment three
At the system that adopts the resource pool scheme, the present invention also provides a kind of method that reduces effect of single plate failure on service, referring to Fig. 3, specifically may further comprise the steps:
Step 201: the various faults that occur according to veneer provide professional influence to veneer, calculate the fault traffic lost rate of every kind of fault;
The computing formula of fault traffic lost rate LF is:
The total professional ability that provides professional ability/veneer to provide of veneer loss is provided fault traffic lost rate LF=fault.
Step 202: the various fault traffic lost rate summation operation to veneer, obtain the single board service loss late, promptly be calculated as follows:
Single board service loss late LB=∑ LFi (1＜i＜M):
Wherein M is the fault sum that occurs on this veneer, and the fault traffic lost rate of i fault correspondence is LFi, the addition of M on the veneer fault traffic lost rate, as the total traffic lost rate LB of this veneer.
When single board service loss late LB surpasses 100%, be 100% with single board service loss late value.That is:
Single board service loss late LB=min (∑ LFi, 100%).
Step 203: be calculated as follows the veneer load factor:
1) when 0≤single board service loss late＜100%:
Veneer load factor=veneer current business amount/(single board service total amount * (1-single board service loss late));
2) when single board service loss late=100%:
Veneer load factor=100%;
Present embodiment has proposed the formula of veneer load factor according to the single board service loss late, has reflected that more accurately veneer externally provides professional ability.
Step 204: the traffic lost rate that compares all veneers in the resource pool, in the load of considering other resource (as CPU), give the veneer distribution service that has minimum veneer load factor in the resource pool, other veneer is distribution service not, thereby has guaranteed the load balancing of each veneer in the resource pool.
No matter be embodiment one, or embodiment three, the situation of a veneer all may appear having only in the system, though there is not redundant veneer, the single board service loss late of calculating can reflect that also veneer externally provides professional ability, and by the single board service loss late, makes the user can understand the situation of veneer operation in time, and make corresponding processing, as reset, change failing single board etc.
Embodiment four
At the system that adopts the resource pool scheme, the present invention also provides a kind of device that reduces effect of single plate failure on service, referring to Fig. 4, specifically comprises:
(1) computing module is used for haveing various fault traffic lost rates according to the various calculation of fault that veneer occurs; Also be used for all fault traffic lost rate summations are obtained the single board service loss late; Also be used for calculating the veneer load factor according to the single board service loss late;
(2) processing module is used to judge whether the veneer load factor is veneer load factor minimum in the system, if then give this veneer distribution service, otherwise do not give this veneer distribution service.
Computing module comprises:
1) fault traffic lost rate arithmetic element is used to be calculated as follows the traffic lost rate that is out of order:
The total professional ability that provides professional ability/veneer to provide of veneer loss is provided fault traffic lost rate=fault;
2) single board service loss late arithmetic element is used for the various fault traffic lost rate summations to veneer, obtains the single board service loss late;
3) veneer load factor arithmetic element is used to be calculated as follows out the veneer load factor:
When the single board service loss late less than 100% the time,
Veneer load factor=veneer current business amount/(single board service total amount * (1-single board service loss late));
When the single board service loss late equaled 100%, getting the veneer load factor was 100%.
More than have only the fault that business is had a direct impact among all embodiment, failure definition traffic lost rate.Business is not had the fault of influence, and as the fault of backup link, not influencing veneer externally provides professional ability, does not then need to analyze, without failure definition traffic lost rate.Business there is the fault of remote-effects,, but reflects the traffic lost rate by the direct traffic affecting fault that it causes also without failure definition traffic lost rate.
Above-described embodiment, the present invention embodiment a kind of more preferably just, the common variation that those skilled in the art carries out in the technical solution of the present invention scope and replacing all should be included in protection scope of the present invention.
Claims (13)
1. a method that reduces effect of single plate failure on service is characterized in that, described method specifically may further comprise the steps:
The various calculation of fault that occur according to veneer have various fault traffic lost rates;
All fault traffic lost rate summations to described veneer obtain the single board service loss late; Wherein, described veneer comprises mainboard and slave board;
Whether switch described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer.
2. the method for reduction effect of single plate failure on service according to claim 1 is characterized in that, described fault traffic lost rate is calculated as follows:
The total professional ability that provides professional ability/veneer to provide of veneer loss is provided fault traffic lost rate=fault.
3. the method for reduction effect of single plate failure on service according to claim 1 is characterized in that, described step of whether switching described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer is specially:
When mainboard traffic lost rate in the described veneer greater than described veneer in during slave board traffic lost rate, switch described mainboard and slave board; Otherwise do not switch.
4. the method for reduction effect of single plate failure on service according to claim 1, it is characterized in that, set in advance a threshold value, described step of whether switching described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer is specially:
When mainboard traffic lost rate in the described veneer greater than described veneer in slave board traffic lost rate, and described mainboard traffic lost rate is switched described mainboard and slave board during also more than or equal to described preset threshold value; Otherwise do not switch.
5. the method for reduction effect of single plate failure on service according to claim 1, it is characterized in that, set in advance a hysteresis, described step of whether switching described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer is specially:
When the difference of slave board traffic lost rate in mainboard traffic lost rate and the veneer in the described veneer during, switch described mainboard and slave board greater than described default hysteresis; Otherwise do not switch.
6. the method for reduction effect of single plate failure on service according to claim 1, it is characterized in that, set in advance a threshold value and a hysteresis, described step of whether switching described mainboard and slave board according to slave board traffic lost rate decision in mainboard traffic lost rate and the veneer in the described veneer is specially:
The difference of slave board traffic lost rate is greater than described default hysteresis in mainboard traffic lost rate and veneer in the described veneer, and described mainboard traffic lost rate is switched described mainboard and slave board during more than or equal to described preset threshold value; Otherwise do not switch.
7. a device that reduces effect of single plate failure on service is characterized in that, described device comprises:
Computing module is used for haveing various fault traffic lost rates according to the various calculation of fault that veneer occurs, and also is used for all the fault traffic lost rate summations to described veneer, obtains the single board service loss late; Wherein, described veneer comprises mainboard and slave board;
Processing module is used for whether switching described mainboard and slave board according to slave board traffic lost rate decision in described veneer mainboard traffic lost rate and the veneer.
8. the device of reduction effect of single plate failure on service according to claim 7 is characterized in that, described computing module comprises:
Fault traffic lost rate arithmetic element is used to be calculated as follows out various fault traffic lost rates:
The total professional ability that provides professional ability/veneer to provide of veneer loss is provided fault traffic lost rate=fault;
Single board service loss late arithmetic element is used for all the fault traffic lost rate summations to described veneer, obtains the single board service loss late, and wherein, described veneer comprises mainboard and slave board.
9. the device of reduction effect of single plate failure on service according to claim 8 is characterized in that, described processing module comprises:
Whether first processing unit is used for judging described veneer mainboard traffic lost rate greater than slave board traffic lost rate in the described veneer, if greater than, then switch described mainboard and slave board; Otherwise do not switch;
Or second processing unit, be used to set in advance a threshold value, when mainboard traffic lost rate in the described veneer greater than described veneer in slave board traffic lost rate, and described mainboard traffic lost rate is switched described mainboard and slave board during also more than or equal to described preset threshold value; Otherwise do not switch;
Or the 3rd processing unit, be used to set in advance a hysteresis, when the difference of slave board traffic lost rate in mainboard traffic lost rate and the veneer in the described veneer during, switch described mainboard and slave board greater than described default hysteresis; Otherwise do not switch;
Or manage the unit everywhere, be used to set in advance a threshold value and a hysteresis, the difference of slave board traffic lost rate is greater than described default hysteresis in mainboard traffic lost rate and veneer in the described veneer, and described mainboard traffic lost rate is switched described mainboard and slave board during more than or equal to described preset threshold value; Otherwise do not switch.
10. a method that reduces effect of single plate failure on service is characterized in that, described method specifically may further comprise the steps:
The various calculation of fault that occur according to veneer have various fault traffic lost rates;
All fault traffic lost rate summations to described veneer obtain the single board service loss late;
Calculate the veneer load factor according to described single board service loss late, judge then whether described veneer load factor is veneer load factor minimum in the system, if then give this veneer distribution service, otherwise do not give this veneer distribution service.
11. the method for reduction effect of single plate failure on service according to claim 10 is characterized in that, described veneer load factor is calculated as follows:
When described single board service loss late less than 100% the time,
Veneer load factor=veneer current business amount/(single board service total amount * (1-single board service loss late));
When described single board service loss late equaled 100%, getting the veneer load factor was 100%.
12. a device that reduces effect of single plate failure on service is characterized in that, described device comprises:
Computing module is used for haveing various fault traffic lost rates according to the various calculation of fault that veneer occurs; Also be used for all fault traffic lost rate summations, obtain the single board service loss late described veneer; Also be used for calculating the veneer load factor according to described single board service loss late;
Processing module is used to judge whether described veneer load factor is veneer load factor minimum in the system, if then give this veneer distribution service, otherwise do not give this veneer distribution service.
13. the device of reduction effect of single plate failure on service according to claim 12 is characterized in that, described computing module comprises:
Veneer load factor arithmetic element is used to be calculated as follows out the veneer load factor:
When described single board service loss late less than 100% the time,
Veneer load factor=veneer current business amount/(single board service total amount * (1-single board service loss late));
When described single board service loss late equaled 100%, getting the veneer load factor was 100%.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2006101383694A CN1949699B (en) | 2006-11-09 | 2006-11-09 | Method and apparatus for reducing single-board fault affect on service |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2006101383694A CN1949699B (en) | 2006-11-09 | 2006-11-09 | Method and apparatus for reducing single-board fault affect on service |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1949699A CN1949699A (en) | 2007-04-18 |
| CN1949699B true CN1949699B (en) | 2010-11-03 |
Family
ID=38019094
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2006101383694A Expired - Fee Related CN1949699B (en) | 2006-11-09 | 2006-11-09 | Method and apparatus for reducing single-board fault affect on service |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN1949699B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101163047B (en) * | 2007-11-23 | 2010-08-25 | 上海华为技术有限公司 | Method and device for implementing rearrangement of main equipment and standby equipment |
| CN101848034B (en) * | 2010-05-05 | 2014-12-17 | 中兴通讯股份有限公司 | Optical path protecting device and method |
| WO2011157145A2 (en) * | 2011-05-31 | 2011-12-22 | 华为技术有限公司 | Main/backup switching method between communication device, communication device, system and service request device |
| WO2011157149A2 (en) * | 2011-05-31 | 2011-12-22 | 华为技术有限公司 | Method, communication device and system, and service request device for main/standby switch between communication devices |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1373427A (en) * | 2001-03-01 | 2002-10-09 | 深圳市中兴通讯股份有限公司 | Device and method for implementing dual system slots |
| CN1529459A (en) * | 2003-10-16 | 2004-09-15 | 港湾网络有限公司 | Main-standby rotation realizing method facing to high-side exchange board |
| CN1710869A (en) * | 2005-07-14 | 2005-12-21 | 广东省电信有限公司研究院 | Method for realizing connected reinforced main-apparatus protection in automatic exchange optical network |
-
2006
- 2006-11-09 CN CN2006101383694A patent/CN1949699B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1373427A (en) * | 2001-03-01 | 2002-10-09 | 深圳市中兴通讯股份有限公司 | Device and method for implementing dual system slots |
| CN1529459A (en) * | 2003-10-16 | 2004-09-15 | 港湾网络有限公司 | Main-standby rotation realizing method facing to high-side exchange board |
| CN1710869A (en) * | 2005-07-14 | 2005-12-21 | 广东省电信有限公司研究院 | Method for realizing connected reinforced main-apparatus protection in automatic exchange optical network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1949699A (en) | 2007-04-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103580902B (en) | Computer information system and dynamic disaster recovery method thereof | |
| Garg et al. | Analysis of preventive maintenance in transactions based software systems | |
| KR101298558B1 (en) | Method and system for switching main/standby boards | |
| CN101227315B (en) | Dynamic state server colony and control method thereof | |
| CN1949699B (en) | Method and apparatus for reducing single-board fault affect on service | |
| CN102244669B (en) | Software upgrading method in stacking equipment and stacking equipment | |
| CN112671928B (en) | Equipment centralized management architecture, load balancing method, electronic equipment and storage medium | |
| GB2508041A (en) | Rack server power management system | |
| CN102035683A (en) | Control method and system for switching of main board and standby board | |
| CN112527567A (en) | System disaster tolerance method, device, equipment and storage medium | |
| CN103647723A (en) | Method and system for monitoring flow | |
| US20070180287A1 (en) | System and method for managing node resets in a cluster | |
| CN114337944A (en) | System-level main/standby redundancy general control method | |
| CN113254205A (en) | Load balancing system, method and device, electronic equipment and storage medium | |
| CN111309132B (en) | Method for multi-gear power supply redundancy of server | |
| JP2007280155A (en) | Reliability improving method in dispersion system | |
| CN101345631B (en) | Management method and system for blade server system exchange plane | |
| CN100395962C (en) | Method and system for equipment switching in communication system | |
| CN116340058A (en) | Master-slave switching method and device | |
| CN101119242B (en) | Communication system cluster method, device and cluster service system applying the same | |
| CN103718536A (en) | System and method for powering redundant components | |
| CN201196776Y (en) | Single plate and system with clock backup | |
| CN114168071A (en) | Distributed cluster capacity expansion method, distributed cluster capacity expansion device and medium | |
| CN103716186B (en) | Artificial telephone traffic system with network fault tolerance capability and method thereof | |
| Sun et al. | A generic availability model for clustered computing systems |
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 | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20101103 Termination date: 20131109 |