CN102480502A - I/O load equilibrium method and I/O server - Google Patents
I/O load equilibrium method and I/O server Download PDFInfo
- Publication number
- CN102480502A CN102480502A CN2010105625580A CN201010562558A CN102480502A CN 102480502 A CN102480502 A CN 102480502A CN 2010105625580 A CN2010105625580 A CN 2010105625580A CN 201010562558 A CN201010562558 A CN 201010562558A CN 102480502 A CN102480502 A CN 102480502A
- Authority
- CN
- China
- Prior art keywords
- load
- server
- clock
- migrated
- value
- 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.)
- Granted
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides an I/O load equilibrium method and an I/O server. The I/O load equilibrium method is applied in the I/O server. A storage system is formed by at least two I/O servers. The method comprises the following steps: calculating a second load value of a second moment according to a prediction algorithm in a first moment; acquiring a load average of the storage system in the second moment; when the second load value is greater than the load average, migrating a data file to be migrated to the target I/O server, wherein the load value of the target I/O server in the second moment is less than the second load value. By using the method in the embodiment of the invention, a load balancing strategy is determined based on the load value in the next moment so that the equilibrium of the system load in the next moment can be realized. Therefore, even if there is network communication time delay right now, the influence of the network communication time delay on the system load equilibrium in the next moment is small. Frequent load migration caused by the network communication time delay can be reduced.
Description
Technical field
The present invention relates to technical field of data storage, relate in particular to a kind of I/O load-balancing method and I/O server.
Background technology
Along with the rapid growth of data message amount, cluster file system, cluster are stored in high-performance calculation, the Internet, video field and are widely used.The cluster storage is the storage mode that data is distributed to each node in the cluster, can the storage solution that possess high I/O bandwidth and flexible extensibility be provided for unstructured data.
In the cluster storage, along with rolling up of I/O number of servers, can cause serious network delay and load unbalanced phenomena, reduce the I/O service quality of cluster storage system, reduce the utilance of storage resources.In order to make full use of cluster storage existing resources, further improve parallel I/O access ability, realize between the I/O server that load balancing just seems to be even more important.
In the prior art; Changing loadization and irregular in most of cluster storages adopts dynamic load balancing method usually, and is concrete; Monitor and collect the load data of every I/O server in real time by a centralized servers; According to the data of collecting, make a strategic decision how to carry out the migration of load then, to realize the load balancing between each I/O server of current time through algorithm.But when storage system was on a grand scale, the traffic between server can be very big; If the whole system load is very high; Tangible network service time-delay phenomenon more can occur, and the network service time-delay centralized servers collection load data can be had influence on, and then making of load-balancing decision can be postponed.For example; If current time has A, two I/O server loads height of B, the load of dynamic load leveling decision migration A, B, but because the network service time-delay has caused the delay of load balancing strategy; When waiting to carry out this load balancing strategy and carrying out load migration; The load of A, B possibly reduce, and when load balancing next time, possibly move the load of just moving away from A, B return again again.Therefore, dynamic load balancing method of the prior art can produce the frequent migration of load between the I/O server when the network service time-delay occurring.
Summary of the invention
The embodiment of the invention provides a kind of I/O load-balancing method and I/O server, can when the network service time-delay occurring, reduce the frequent migration of load between the I/O server.
In order to solve the problems of the technologies described above, the technical scheme of the embodiment of the invention is following:
The embodiment of the invention provides a kind of I/O load-balancing method, and said method is used in the I/O server, constitutes a storage system with at least two said I/O servers, comprising:
Inscribed second load value that calculated for second moment according to a prediction algorithm at first o'clock;
Obtain the load average value that said storage system was inscribed at said second o'clock;
When said second load value during greater than said load average value, data file to be migrated is migrated to target I/O server, said target I/O server at the said load value of inscribing in second o'clock less than said second load value.
Further, also comprise:
Obtain first load value inscribed at said first o'clock;
When said second load value during, carry out the said step of obtaining the load average value that said storage system inscribed at said second o'clock again greater than said first load value.
Further, also comprise:
Obtain the load balancing factor that said storage system was inscribed at said second o'clock;
When the said load balancing factor during, carry out the said step that data file to be migrated is migrated to target I/O server again less than load balancing factor threshold value.
Further, said I/O server obtains the load average value that the place storage system was inscribed at said second o'clock, comprising:
Other I/O server to said storage system sends the request of collecting the said load value of inscribing at second o'clock;
Receive said other I/O server feedback at the said load value of inscribing in second o'clock;
Calculate load average value that said storage system at said second o'clock inscribe according to said other I/O server at said load value of inscribing and said second load value at second o'clock.
Further, said data file to be migrated is migrated to target I/O server, comprising:
Select data file to be migrated;
Send migration request to said target I/O server;
Receive the migration response that said target I/O server returns;
Said data file to be migrated is sent to said target I/O server, and the local data file said to be migrated of deletion.
Further, the data file that said selection is to be migrated comprises:
Calculate the load value of each data file of storing and the ratio of data volume;
The data file of selecting the maximum predetermined quantity of ratio is as data file to be migrated.
Further, also comprise:
Confirm a said target I/O server according to a predetermined policy at other I/O servers of said storage system;
To be said target I/O server be not more than said I/O server at the said load value second o'clock inscribed in the load value sum of said load value of inscribing and said data file to be migrated at second o'clock to said predetermined policy.
The embodiment of the invention also provides a kind of I/O server, is applied to comprise in the storage system of at least two said I/O servers formations:
The load estimation unit is used for inscribing at first o'clock according to a prediction algorithm and calculates second constantly second load value;
The load acquiring unit is used to obtain the load average value that said storage system was inscribed at said second o'clock;
The load migration unit is used for when said second load value during greater than said load average value data file to be migrated being migrated to target I/O server, said target I/O server at the said load value of inscribing in second o'clock less than said second load value.
Further, also comprise:
The load calculation unit is used to calculate first load value of inscribing at first o'clock;
Said load acquiring unit also is used for obtaining the load average value that said storage system was inscribed at said second o'clock again when said second load value during greater than said first load value.
Further, also comprise:
The balance factor acquiring unit is used to obtain the load balancing factor that said storage system was inscribed at said second o'clock;
Said load migration unit also is used for again data file to be migrated being migrated to target I/O server when the said load balancing factor during less than load balancing factor threshold value.
Further, said load acquiring unit comprises:
The load requests subelement is used for sending the request of collecting the said load value of inscribing at second o'clock to other I/O server of said storage system;
Load receives subelement, be used to receive said other I/O server feedback at the said load value of inscribing in second o'clock;
Computation subunit is used for calculating load average value that said storage system at said second o'clock inscribe according to said other I/O server at said load value of inscribing and said second load value at second o'clock.
Further, said load migration unit comprises:
The chooser unit is used to select data file to be migrated;
The migration request subelement is used for sending migration request to said target I/O server;
Response receives subelement, is used to receive the migration response that said target I/O server sends;
The migration subelement is used for said data file to be migrated is sent to said target I/O server, and the local data file said to be migrated of deletion.
Further, said chooser unit comprises:
The ratio calculation subelement is used to calculate load value and the ratio of data volume of each data file of storage;
File chooser unit, the data file that is used to select the maximum predetermined quantity of ratio is as data file to be migrated.
Further, said load migration unit also comprises:
Confirm subelement, be used for confirming a said target I/O server at other I/O servers of said storage system according to a predetermined policy; To be said target I/O server be not more than said I/O server at the said load value second o'clock inscribed in the load value sum of said load value of inscribing and said data file to be migrated at second o'clock to said predetermined policy.
Next load value constantly of I/O server prediction in the embodiment of the invention, and obtain the load average value of system, and then carry out the comparison of load value when comparative result satisfies condition, is initiated load migration again, the load balancing of etching system when satisfying next.Embodiment of the invention method is through confirming the load balancing strategy based on next load value constantly; Realized equilibrium to next system load constantly; So even there is the network service time-delay in current time; The load balancing of etching system does not produce considerable influence in the time of can be to next yet, thereby has reduced the load frequent migration that causes because of the network service time-delay.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art; To do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art below; Obviously, the accompanying drawing in describing below only is some embodiments of the present invention, for those of ordinary skills; Under the prerequisite of not paying creative work property, can also obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the flow chart of a kind of I/O load-balancing method of the embodiment of the invention;
Fig. 2 is the flow chart of the another kind of I/O load-balancing method of the embodiment of the invention;
Fig. 3 is the method flow diagram of load migration in the another kind of I/O load-balancing method of the embodiment of the invention;
Fig. 4 is the structural representation of a kind of I/O server of the embodiment of the invention;
Fig. 5 is the structural representation of the another kind of I/O server of the embodiment of the invention;
Fig. 6 is the structural representation of a kind of storage system of the embodiment of the invention.
Embodiment
In order to make those skilled in the art can further understand characteristic of the present invention and technology contents, see also followingly about detailed description of the present invention and accompanying drawing, accompanying drawing only provides reference and explanation, is not to be used for limiting the present invention.
Dynamic load balancing method of the prior art is because the load data of the I/O server that need collect according to current time; Equilibrium is carried out in load between each I/O server of current time, so, in case network communication delay will cause obtaining delaying of load data; And then can postpone making of load balancing strategy; Because the dynamic change of load, this load balancing strategy possibly not be suitable for the loading condition of each I/O server of time-delay back, carry out this load balancing strategy again and then cause inaccurate load migration possibly; Even the load that can increase the weight of system is unbalanced, and then can produce the frequent migration of load.
Based on this; The embodiment of the invention has proposed a kind of I/O load-balancing method and I/O server, next load value constantly of I/O server prediction, and obtain the load average value of system; And then carry out the comparison of each load value; When comparative result satisfies condition, initiate load migration again, the load balancing of etching system when satisfying next.Embodiment of the invention method is through confirming the load balancing strategy based on next load value constantly; Realized equilibrium to next system load constantly; So even there is the network service time-delay in current time; The load balancing of etching system does not produce considerable influence in the time of can be to next yet, thereby has reduced the load frequent migration that causes because of the network service time-delay.
Below in conjunction with accompanying drawing and embodiment, technical scheme of the present invention is described.
Referring to Fig. 1, be the flow chart of a kind of I/O load-balancing method of the embodiment of the invention.
This method is applied in the I/O server, and at least two I/O servers constitute a storage system, and this method can comprise:
In the present embodiment; The I/O server is predicted at next moment (second constantly) possible load value own according to the prediction algorithm that obtains based on certain forecast model; I.e. this second load value prediction load value that is this I/O server is not the actual value of inscribing at second o'clock yet.Wherein, first is merely two time points distinguishing front and back constantly with second constantly, is not to refer in particular to specifically point sometime.Two time intervals between constantly can be set as required, and for example very fast if the system load situation changes, then setting-up time is shorter at interval, and is slower if loading condition changes, but then setting-up time is long at interval etc.Prediction algorithm can be that for example linear prediction auto-correlation (Levinson-Durbin) algorithm, red pond information criterion, autoregression AR forecast model etc. repeat no more here according to existing model and algorithm.Load value can be meant the IO throughput of I/O server etc., specifically can set according to the parameter that will realize load balancing.
The I/O server can be real-time when carrying out this step, also can be wake up in the cycle also promptly periodically carry out this step, this cycle also can be set as required, repeats no more here.
Step 102 is obtained the load average value that storage system was inscribed at second o'clock.
The I/O server inscribed in first o'clock obtain the load value of inscribing at second o'clock after; Further obtain the load average value of inscribing this I/O server place storage system at second o'clock; Concrete acquisition methods can be to calculate behind the load value when other I/O server requests the second, inscribed of this I/O server to obtain; Also can directly send, specifically see also the description of subsequent implementation example by other I/O server.The load average value of inscribing in this second o'clock also is predicted value.
Step 103 when second load value during greater than the load average value, migrates to target I/O server with data file to be migrated.
When second load value of I/O server during greater than the load average value of storage system; Explain this I/O server at second constantly the loading condition with respect to other I/O server in the storage system; To be in high load condition; To reach the load balancing of storage system in second moment, need the partial document data in this I/O server is migrated in the relatively low I/O server of load.At this moment, this I/O server data file that this locality is to be migrated migrates in the target I/O server, and the load value that this target I/O server was inscribed at second o'clock is less than second load value.
Certainly in this step, can also increase other qualifications, with under the prerequisite that realizes second moment storage system load balancing, reduce load migration as far as possible.Specifically see also the subsequent implementation example.
In the present embodiment, every I/O server in the storage system all moves above-mentioned equalization methods, and the structure of this independent distribution formula can be avoided the problem of the bottleneck and the Single Point of Faliure of traditional centralized management method; This load-balancing algorithm can reduce the mutual dependence between the I/O server, lets workload as far as possible on book server, has only the server ability needs that satisfy certain condition further to collect necessary information.The embodiment of the invention is through confirming the load balancing strategy based on next load value constantly; Realized equilibrium to next system load constantly; So even there is the network service time-delay in current time; The load balancing of etching system does not produce considerable influence in the time of can be to next yet, thereby has reduced the load frequent migration that causes because of the network service time-delay.
Referring to Fig. 2, be the flow chart of the another kind of I/O load-balancing method of the embodiment of the invention.
This method can comprise:
The I/O server can adopt the computational methods of prior art, calculates the load value a that inscribed at first o'clock, and this first moment also is a current time.
Predict the load value b that inscribed at second o'clock according to prediction algorithm then.Step 101 in this step and the previous embodiment is similar, repeats no more here.
Step 203 as b during greater than a, is obtained the load average value c that I/O server place storage system was inscribed at second o'clock.
During load value a that the load value b that inscribes second time when the I/O server inscribes greater than first time; The load that this I/O server is described will increase; And then obtain the load average value c that inscribed that this I/O server belongs to storage system at second o'clock, to determine whether need to carry out load migration to realize the load balancing of storage system.
If b is smaller or equal to c; Explain that also there is the situation of load rise in interior other I/O server of storage system; And it is more to raise than this I/O server load, and the loading condition of this I/O server still is in average level or subaverage, and then this I/O server need not to carry out load migration.If it is comparatively serious that b, explains that this I/O server load raises greater than c, change next step over to and further judge whether to carry out load migration.
Wherein, the load average value c that inscribed in second of storage system o'clock can obtain through following steps:
This I/O server sends the request of collecting the load value of inscribing at second o'clock to other I/O server of place storage system;
This I/O server receives the load value of inscribing at second o'clock that other I/O server sends;
Load value that this I/O server was inscribed at second o'clock according to other I/O server and second load value of this I/O server calculate the load average value that storage system was inscribed at second o'clock.
In other embodiments, also can be that the load value that this I/O server will own be inscribed at second o'clock is sent to other a certain I/O server, behind other I/O server calculating acquisition mean value, be sent to this I/O server.
In the present embodiment, the I/O server can also further obtain the load balancing factor that storage system was inscribed at second o'clock, and this load balancing factor can embody storage system in the degree of whole system load balancing sometime.If this load balancing factor explains then that less than the preset load balancing factor threshold value load of storage system is unbalance, need carry out load migration to realize load balancing.
Wherein, this load balancing factor threshold value is for weighing the lower limit of this storage system load balancing, and less than this lower limit explanation load imbalance, concrete numerical value can be set as required, does not limit here.
Load average value c that the method that the I/O server obtains the storage system load balancing factor can be based on storage system calculates and obtains; The load average value of storage system that concrete can be accounts for the percentage of storage system maximum load value, and this maximum load value is to set according to the memory property of storage system.This load balancing factor is big more, explains that the load of storage system is more balanced.The computational methods of load balancing require effectively simple, do not increase the extra load of I/O server under the high capacity as far as possible.After the reference conditions as migration after the adding load balancing factor, can improve the availability of load migration decision-making, avoid the action of some invalid load migrations.
Step 205, when b greater than c, and the load balancing factor of storage system is during less than load balancing factor threshold value, the I/O server migrates to target I/O server with data file to be migrated.
When the load average value of this I/O server second load value constantly greater than storage system; And the load balancing factor of storage system is less than load balancing factor threshold value; Explain that then this I/O server need carry out load migration, to reduce the load value of self, with the load of balanced storage system.Obviously, this I/O server carry out the load value of the target I/O server of load migration at least need be less than the load value of this I/O server.
Wherein, concrete load migration process can be as shown in Figure 3, comprising:
In the present embodiment, after load migration is carried out in the decision of I/O server, need to select data file to be migrated, this selection course can be: at first, calculate the load value of each data file of storing and the ratio of data volume.Then, the data file of selecting the maximum some of ratio is as data file to be migrated.
Wherein, Data file load value and data volume ratio are big more; Explain that the load value that this data file accounts for is big more in this I/O server, can preferentially select to move this data file, then can realize moving the load value that a small amount of this class data file can reduce this I/O server fast.The data file of selecting the maximum some of ratio is as data file to be migrated, and wherein, this some can be made an appointment, and in the present embodiment, can select maximum preceding ten data files of ratio as data file to be migrated.
In the present embodiment, this predetermined policy can be:
The load value sum of load value that target I/O server was inscribed at second o'clock and data file to be migrated is not more than the load value that this I/O server was inscribed at second o'clock; Perhaps, the load value sum of the load value inscribed at second o'clock of target I/O server and data file to be migrated is not more than load value poor of load value that this I/O server inscribed at second o'clock and data file to be migrated.Certainly, in other embodiments, can also set the condition of other select target I/O server, enumerate no longer one by one here according to the loading condition of storage system.
Adopting such strategy is in order to judge when certain load during from the I/O server migration of high capacity to low load targets I/O server; The load of target I/O server can not be higher than I/O server second load value constantly; Could eliminate high capacity I/O server like this, the load distribution situation of the whole storage system of efficient balance.
This step can be carried out with last step or the adjustment order simultaneously.
After this I/O server selected data file and target I/O server to be migrated, promptly obtain the attribute of data file to be migrated, and send migration request to target I/O server, can comprise the attribute of this data file to be migrated in this request.
After target I/O server receives migration request, can set up new file attribute in this locality, send the migration response to the I/O server then according to the attribute of the data file to be migrated that comprises in this request.
The I/O server can be sent to target I/O server with data file to be migrated after receiving the migration response, then this data file of this locality is deleted.After migration was accomplished, the I/O server can also further notify meta data server to upgrade the distributed intelligence of this data file.
Embodiment of the invention method has realized the equilibrium to next system load constantly, and after adding the load balancing factor, can improve the availability of load migration decision-making, avoids the action of some invalid load migrations.So even there is network service time-delay in current time, the load balancing of etching system does not produce considerable influence in the time of can be to next yet, thereby has reduced the load frequent migration that causes because of the network service time-delay; On the other hand, the situation of whether carrying out load balancing is screened, more reduced the frequent migration of load through each load value relatively.
Referring to Fig. 4, be the structural representation of a kind of I/O server of the embodiment of the invention.
In the present embodiment, in the storage system that this I/O server is applied at least to be made up of two I/O servers, this I/O server can comprise:
Load estimation unit 401 is used for inscribing at first o'clock according to a prediction algorithm and calculates second constantly second load value.
Load acquiring unit 402 is used to obtain the load average value that storage system was inscribed at second o'clock.
Load migration unit 403 is used for when second load value during greater than the load average value data file to be migrated being migrated to target I/O server, and the load value that target I/O server was inscribed at second o'clock is less than second load value.
Load estimation unit 401 is predicted at next possible load value of the moment own based on certain prediction algorithm.Prediction algorithm can be that for example linear prediction auto-correlation (Levinson-Durbin) algorithm, red pond information criterion, autoregression AR forecast model etc. repeat no more here according to existing model and/or algorithm.Load acquiring unit 402 further obtains the load average value of inscribing this I/O server place storage system at second o'clock; Concrete acquisition methods can be to calculate behind the load value when other I/O server requests the second, inscribed of this I/O server to obtain, and also can directly be sent by other I/O server.When second load value of I/O server during greater than the load average value of storage system; Explain this I/O server at second constantly the loading condition with respect to other I/O server in the storage system; Be in high load condition; To reach the load balancing of storage system in second moment, need the partial document data in this I/O server is migrated in the relatively low I/O server of load.At this moment, load migration unit 403 data file that this locality is to be migrated migrates in the target I/O server, realizes the equilibrium of storage system load.
The embodiment of the invention has realized the equilibrium to next system load constantly through said units; So even there is the network service time-delay in current time; The load balancing of etching system does not produce considerable influence in the time of can be to next yet, thereby has reduced the load frequent migration that causes because of the network service time-delay.
Referring to Fig. 5, be the structural representation of the another kind of I/O server of the embodiment of the invention.
This I/O server can also comprise load calculation unit 504 and balance factor acquiring unit 505 except can comprising load estimation unit 501, load acquiring unit 502, load migration unit 503.
Wherein, load estimation unit 501 is similar with the load estimation unit 401 in the previous embodiment, repeats no more here.
Load calculation unit 504 is used to calculate first load value of inscribing at first o'clock.
Load acquiring unit 502 can also be used for obtaining the load average value that storage system was inscribed at second o'clock again when second load value during greater than first load value.
Balance factor acquiring unit 505 is used to obtain the load balancing factor that storage system was inscribed at second o'clock.
Load migration unit 503 can also be used for again data file to be migrated being migrated to target I/O server when the load balancing factor during less than load balancing factor threshold value.
Concrete, load acquiring unit 502 can comprise again:
Load requests subelement 5021 is used for sending the request of collecting the load value of inscribing at second o'clock to other I/O server of storage system.
Load receives subelement 5022, be used to receive other I/O server feedback at the said load value of inscribing in second o'clock.
Computation subunit 5023, the load value and second load value that are used for inscribing at second o'clock according to other I/O server calculate the load average value that storage system was inscribed at second o'clock.
Load migration unit 503 can comprise again:
Chooser unit 5031 is used to select data file to be migrated.Chooser unit 5031 can also comprise ratio calculation subelement and file chooser unit.Wherein, the ratio calculation subelement is used to calculate load value and the ratio of data volume of each data file of storage; The data file that file chooser unit is used to select the maximum predetermined quantity of ratio is as data file to be migrated.
Confirm subelement 5032, be used for confirming a target I/O server at other I/O servers of storage system according to a predetermined policy; The load value sum of load value that predetermined policy can be inscribed at second o'clock for target I/O server and data file to be migrated is not more than the load value that the I/O server was inscribed at second o'clock.
Migration request subelement 5033 is used for sending migration request to target I/O server.
Response receives subelement 5034, is used for the migration response that receiving target I/O server sends.
Migration subelement 5035 is used for data file to be migrated is sent to target I/O server, and the local data file to be migrated of deletion.
501 predictions of load estimation unit obtain second load value that the I/O server was inscribed at second o'clock; Load calculation unit 504 calculates and obtains the I/O server at first o'clock after first load value of inscribing; Through comparing; If second load value is greater than first load value; Then load acquiring unit 502 receives subelement 5022 and computation subunit 5023 through load requests subelement 5021, load, obtains the load average value that the storage system at this I/O server place was inscribed at second o'clock, and balance factor acquiring unit 505 obtains the load balancing factor that storage system was inscribed at second o'clock.When second load value of this I/O server greater than the load average value; And the load balancing factor of storage system is during less than preset threshold value, load migration unit 503 through chooser unit 5031, confirm that subelement 5032, migration request subelement 5033, response receive subelement 5034 and migration subelement 5035 migrates to target I/O server with data file to be migrated.The load value sum of load value that this target I/O server was inscribed at second o'clock and data file to be migrated is not more than the load value that this server of I/O was inscribed at second o'clock.
The embodiment of the invention is confirmed the load balancing strategy through said units based on next load value constantly; Realized equilibrium to next system load constantly; So even there is the network service time-delay in current time; The load balancing of etching system does not produce considerable influence in the time of can be to next yet, thereby has reduced the load frequent migration that causes because of the network service time-delay; On the other hand, the situation of whether carrying out load balancing is screened, more reduced the frequent migration of load through each load value of a plurality of unit comparison.
Referring to Fig. 6, be the structural representation of a kind of storage system of the embodiment of the invention.
Comprise at least two I/O servers in this storage system, present embodiment is an example with three I/O servers 601,602,603.
Every I/O server all is used for inscribing at first o'clock according to a prediction algorithm and calculates second constantly second load value; Obtain the load average value that storage system was inscribed at second o'clock; When second load value during greater than the load average value, data file to be migrated is migrated to target I/O server, the load value that target I/O server was inscribed at second o'clock is less than second load value.
Every I/O server can also at first obtain first o'clock first load value of inscribing, and when second load value during greater than first load value, obtains the step of the load average value that storage system inscribed at second o'clock again.Every I/O server also can also at first obtain the load balancing factor that storage system was inscribed at second o'clock, when the load balancing factor during less than load balancing factor threshold value, data file to be migrated is migrated to the step of target I/O server again.
In the present embodiment; I/O server 601,602,603 calculates the load value under the own current time (also being first moment) respectively; And calculate the load value of inscribing at second o'clock according to prediction algorithm; Ask other I/O server second load value constantly then respectively, calculate the load average value that obtains storage system, and calculate the load balancing factor that obtains to inscribe storage system at second o'clock.Suppose that load value that the I/O server inscribed is greater than the load value of inscribing at 601 second o'clock at first o'clock; And greater than the load average value; And the load balancing factor of inscribing storage system at second o'clock is less than load balancing factor threshold value; Then I/O server 601 selects the maximum data file of the ratio of load value and data volume as data file to be migrated in the data file of this locality; The load value sum of load value of in I/O server 602,603, selecting to inscribe in second o'clock and data file to be migrated is not more than of load value that I/O server 601 inscribed at second o'clock, as target I/O server, for example is I/O server 602.I/O server 601 sends migration request to I/O server 602; After waiting to receive the migration response that I/O server 602 returns; Data file to be migrated is sent to I/O server 602; And the local data file to be migrated of deletion, accomplish data file from the migration of the I/O server 601 of high capacity, thereby realize the load balancing in the storage system to the I/O server 602 of low load.
In the present embodiment, storage system is through all moving above-mentioned equalization methods on every I/O server, and the structure of this independent distribution formula can be avoided the problem of the bottleneck and the Single Point of Faliure of traditional centralized management method; This load-balancing algorithm can reduce the mutual trust between the I/O server, lets workload as far as possible on book server, has only the server ability needs that satisfy certain condition further to collect necessary information.The embodiment of the invention is confirmed the load balancing strategy through every I/O server based on next load value constantly; Realized equilibrium to next system load constantly; So even there is the network service time-delay in current time; The load balancing of etching system does not produce considerable influence in the time of can be to next yet, thereby has reduced the load frequent migration that causes because of the network service time-delay.
More than the concrete implementation procedure of each unit, subelement and system repeats no more please with reference to the corresponding description among the preceding method embodiment here in the device.
Above-described embodiment of the present invention does not constitute the qualification to protection range of the present invention.Any modification of within spirit of the present invention and principle, being done, be equal to replacement and improvement etc., all should be included within the claim protection range of the present invention.
Claims (14)
1. I/O load-balancing method, said method is used in the I/O server, constitutes a storage system with at least two said I/O servers, it is characterized in that, comprising:
Inscribed second load value that calculated for second moment according to a prediction algorithm at first o'clock;
Obtain the load average value that said storage system was inscribed at said second o'clock;
When said second load value during greater than said load average value, data file to be migrated is migrated to target I/O server, said target I/O server at the said load value of inscribing in second o'clock less than said second load value.
2. method according to claim 1 is characterized in that, also comprises:
Obtain first load value inscribed at said first o'clock;
When said second load value during, carry out the said step of obtaining the load average value that said storage system inscribed at said second o'clock again greater than said first load value.
3. method according to claim 1 is characterized in that, also comprises:
Obtain the load balancing factor that said storage system was inscribed at said second o'clock;
When the said load balancing factor during, carry out the said step that data file to be migrated is migrated to target I/O server again less than load balancing factor threshold value.
4. method according to claim 1 is characterized in that, said I/O server obtains the load average value that the place storage system was inscribed at said second o'clock, comprising:
Other I/O server to said storage system sends the request of collecting the said load value of inscribing at second o'clock;
Receive said other I/O server feedback at the said load value of inscribing in second o'clock;
Calculate load average value that said storage system at said second o'clock inscribe according to said other I/O server at said load value of inscribing and said second load value at second o'clock.
5. method according to claim 1 is characterized in that, said data file to be migrated is migrated to target I/O server, comprising:
Select data file to be migrated;
Send migration request to said target I/O server;
Receive the migration response that said target I/O server returns;
Said data file to be migrated is sent to said target I/O server, and the local data file said to be migrated of deletion.
6. method according to claim 5 is characterized in that, the data file that said selection is to be migrated comprises:
Calculate the load value of each data file of storing and the ratio of data volume;
The data file of selecting the maximum predetermined quantity of ratio is as data file to be migrated.
7. method according to claim 6 is characterized in that, also comprises:
Confirm a said target I/O server according to a predetermined policy at other I/O servers of said storage system;
To be said target I/O server be not more than said I/O server at the said load value second o'clock inscribed in the load value sum of said load value of inscribing and said data file to be migrated at second o'clock to said predetermined policy.
8. an I/O server is applied to it is characterized in that in the storage system of at least two said I/O servers formations, comprising:
The load estimation unit is used for inscribing at first o'clock according to a prediction algorithm and calculates second constantly second load value;
The load acquiring unit is used to obtain the load average value that said storage system was inscribed at said second o'clock;
The load migration unit is used for when said second load value during greater than said load average value data file to be migrated being migrated to target I/O server, said target I/O server at the said load value of inscribing in second o'clock less than said second load value.
9. I/O server according to claim 8 is characterized in that, also comprises:
The load calculation unit is used to calculate first load value of inscribing at first o'clock;
Said load acquiring unit also is used for obtaining the load average value that said storage system was inscribed at said second o'clock again when said second load value during greater than said first load value.
10. I/O server according to claim 8 is characterized in that, also comprises:
The balance factor acquiring unit is used to obtain the load balancing factor that said storage system was inscribed at said second o'clock;
Said load migration unit also is used for again data file to be migrated being migrated to target I/O server when the said load balancing factor during less than load balancing factor threshold value.
11. I/O server according to claim 8 is characterized in that, said load acquiring unit comprises:
The load requests subelement is used for sending the request of collecting the said load value of inscribing at second o'clock to other I/O server of said storage system;
Load receives subelement, be used to receive said other I/O server feedback at the said load value of inscribing in second o'clock;
Computation subunit is used for calculating load average value that said storage system at said second o'clock inscribe according to said other I/O server at said load value of inscribing and said second load value at second o'clock.
12. I/O server according to claim 8 is characterized in that, said load migration unit comprises:
The chooser unit is used to select data file to be migrated;
The migration request subelement is used for sending migration request to said target I/O server;
Response receives subelement, is used to receive the migration response that said target I/O server sends;
The migration subelement is used for said data file to be migrated is sent to said target I/O server, and the local data file said to be migrated of deletion.
13. I/O server according to claim 12 is characterized in that, said chooser unit comprises:
The ratio calculation subelement is used to calculate load value and the ratio of data volume of each data file of storage;
File chooser unit, the data file that is used to select the maximum predetermined quantity of ratio is as data file to be migrated.
14. I/O server according to claim 13 is characterized in that, said load migration unit also comprises:
Confirm subelement, be used for confirming a said target I/O server at other I/O servers of said storage system according to a predetermined policy; To be said target I/O server be not more than said I/O server at the said load value second o'clock inscribed in the load value sum of said load value of inscribing and said data file to be migrated at second o'clock to said predetermined policy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010562558.0A CN102480502B (en) | 2010-11-26 | 2010-11-26 | I/O load equilibrium method and I/O server |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010562558.0A CN102480502B (en) | 2010-11-26 | 2010-11-26 | I/O load equilibrium method and I/O server |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102480502A true CN102480502A (en) | 2012-05-30 |
| CN102480502B CN102480502B (en) | 2014-10-01 |
Family
ID=46092982
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010562558.0A Active CN102480502B (en) | 2010-11-26 | 2010-11-26 | I/O load equilibrium method and I/O server |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102480502B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102968281A (en) * | 2012-11-26 | 2013-03-13 | 华为技术有限公司 | Data migration method and device |
| CN103918239A (en) * | 2012-09-28 | 2014-07-09 | 华为技术有限公司 | Load balancing method, device, system and computer readable medium |
| CN104468674A (en) * | 2013-09-25 | 2015-03-25 | 南京中兴新软件有限责任公司 | Data migration method and device |
| CN104573020A (en) * | 2015-01-12 | 2015-04-29 | 浪潮电子信息产业股份有限公司 | Automatic data migrating and optimizing method in hierarchical storage system |
| CN107783720A (en) * | 2016-08-24 | 2018-03-09 | 深圳市深信服电子科技有限公司 | A kind of data balancing method and storage device |
| CN114143326A (en) * | 2021-12-08 | 2022-03-04 | 深圳前海微众银行股份有限公司 | Load adjustment method, management node, and storage medium |
| CN114448897A (en) * | 2021-12-29 | 2022-05-06 | 天翼云科技有限公司 | Target device migration method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1512729A (en) * | 2002-12-31 | 2004-07-14 | 联想(北京)有限公司 | Method for network equipment self adaption load equalization |
| CN101593133A (en) * | 2009-06-29 | 2009-12-02 | 北京航空航天大学 | Load balancing of resources of virtual machine method and device |
| CN101695050A (en) * | 2009-10-19 | 2010-04-14 | 浪潮电子信息产业股份有限公司 | Dynamic load balancing method based on self-adapting prediction of network flow |
| CN101820384A (en) * | 2010-02-05 | 2010-09-01 | 浪潮(北京)电子信息产业有限公司 | Method and device for dynamically distributing cluster services |
-
2010
- 2010-11-26 CN CN201010562558.0A patent/CN102480502B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1512729A (en) * | 2002-12-31 | 2004-07-14 | 联想(北京)有限公司 | Method for network equipment self adaption load equalization |
| CN101593133A (en) * | 2009-06-29 | 2009-12-02 | 北京航空航天大学 | Load balancing of resources of virtual machine method and device |
| CN101695050A (en) * | 2009-10-19 | 2010-04-14 | 浪潮电子信息产业股份有限公司 | Dynamic load balancing method based on self-adapting prediction of network flow |
| CN101820384A (en) * | 2010-02-05 | 2010-09-01 | 浪潮(北京)电子信息产业有限公司 | Method and device for dynamically distributing cluster services |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103918239A (en) * | 2012-09-28 | 2014-07-09 | 华为技术有限公司 | Load balancing method, device, system and computer readable medium |
| CN102968281A (en) * | 2012-11-26 | 2013-03-13 | 华为技术有限公司 | Data migration method and device |
| CN104468674A (en) * | 2013-09-25 | 2015-03-25 | 南京中兴新软件有限责任公司 | Data migration method and device |
| CN104573020A (en) * | 2015-01-12 | 2015-04-29 | 浪潮电子信息产业股份有限公司 | Automatic data migrating and optimizing method in hierarchical storage system |
| CN104573020B (en) * | 2015-01-12 | 2017-11-10 | 浪潮电子信息产业股份有限公司 | Data Autonomic Migration Framework optimization method in a kind of hierarchical stor |
| CN107783720A (en) * | 2016-08-24 | 2018-03-09 | 深圳市深信服电子科技有限公司 | A kind of data balancing method and storage device |
| CN114143326A (en) * | 2021-12-08 | 2022-03-04 | 深圳前海微众银行股份有限公司 | Load adjustment method, management node, and storage medium |
| CN114448897A (en) * | 2021-12-29 | 2022-05-06 | 天翼云科技有限公司 | Target device migration method and device |
| CN114448897B (en) * | 2021-12-29 | 2024-01-02 | 天翼云科技有限公司 | Target migration method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102480502B (en) | 2014-10-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210149737A1 (en) | Method for fast scheduling for balanced resource allocation in distributed and collaborative container platform environment | |
| CN102480502B (en) | I/O load equilibrium method and I/O server | |
| CN107562531B (en) | Data equalization method and device | |
| CN102055650B (en) | Load balance method and system and management server | |
| CN110134495B (en) | Container cross-host online migration method, storage medium and terminal equipment | |
| CN107273185B (en) | Load balancing control method based on virtual machine | |
| CN108182105B (en) | Local dynamic migration method and control system based on Docker container technology | |
| CN107612771B (en) | SDN network load balancing method based on dynamic migration | |
| CN105025053A (en) | Distributed file upload method based on cloud storage technology and system | |
| CN106648456B (en) | Dynamic copies file access method based on user's amount of access and forecasting mechanism | |
| CN103401947A (en) | Method and device for allocating tasks to multiple servers | |
| KR102245341B1 (en) | Method for apply predictive model for workload distribution within the cloud edge | |
| CN102299959A (en) | Load balance realizing method of database cluster system and device | |
| CN105491150A (en) | Load balance processing method based on time sequence and system | |
| US10250673B1 (en) | Storage workload management using redirected messages | |
| CN102984280A (en) | Data backup system and method for social cloud storage network application | |
| CN108255427B (en) | Data storage and dynamic migration method and device | |
| CN109196807B (en) | Network node and method of operating a network node for resource distribution | |
| CN103631894A (en) | Dynamic copy management method based on HDFS | |
| JP2016024500A (en) | Distributed processing program, distributed processing management device, and distributed processing method | |
| CN108200156A (en) | The dynamic load balancing method of distributed file system under a kind of cloud environment | |
| CN105302830A (en) | Map tile caching method and apparatus | |
| EP4170491A1 (en) | Resource scheduling method and apparatus, electronic device, and computer-readable storage medium | |
| CN103918239A (en) | Load balancing method, device, system and computer readable medium | |
| CN105095495A (en) | Distributed file system cache management method 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 |