CN107391633A

CN107391633A - Data-base cluster Automatic Optimal processing method, device and server

Info

Publication number: CN107391633A
Application number: CN201710556177.3A
Authority: CN
Inventors: 李丹
Original assignee: Beijing Qihoo Technology Co Ltd
Current assignee: Beijing Qihoo Technology Co Ltd
Priority date: 2017-06-30
Filing date: 2017-06-30
Publication date: 2017-11-24

Abstract

The invention discloses a kind of data-base cluster Automatic Optimal processing method, device, server and computer-readable storage medium, wherein, this method includes：The status information of monitoring data storehouse cluster, gather the state index characteristic information of data-base cluster corresponding with preset state index；According to the state index characteristic information of data-base cluster, the operation conditions of analytical database cluster；Corresponding troubleshooting measure and/or optimization processing measure, which are triggered, for operation conditions carries out Automatic Optimal processing.According to scheme provided by the invention, the different operation conditions of data-base cluster can be directed to and trigger corresponding troubleshooting measure and/or optimization processing measure progress Automatic Optimal processing, compared with prior art, excessive human resources need not be put into, reduce human cost, and the various failures and problem to be optimized that database occurs can be found in time, these problems are automatically processed, improve the performance and service quality of database.

Description

Data-base cluster Automatic Optimal processing method, device and server

Technical field

The present invention relates to database processing field, and in particular to a kind of data-base cluster Automatic Optimal processing method, device, Server and computer-readable storage medium.

Background technology

Data-base cluster is exactly using at least two or more database servers, forms a virtual centralized database Logical image, as single database system, transparent data, services are provided to client.Carried in data-base cluster to client During for service, with the development of internet, the business datum amount of enterprise's storage is also sharp increase, in such case Down, it is necessary to improve constantly the performance of data-base cluster and ensure that data-base cluster externally provides service quality and keeps stable.

In the prior art, the problem and failure problems to be optimized occurred for data-base cluster, mainly by manually doing Pre- mode is adjusted to irrational parameter in data-base cluster, or pass through caused by artificial detection and processing therefore Barrier, this mode is when solving the above problems, it is impossible to finds irrational setting or failure in time, it is impossible to targetedly handle Problem, causing processing procedure, time-consuming, and consumption resource is more, and the cost input of manual intervention is big.

The content of the invention

In view of the above problems, it is proposed that the present invention so as to provide one kind overcome above mentioned problem or at least in part solve on State data-base cluster Automatic Optimal processing method, device, server and the computer-readable storage medium of problem.

According to an aspect of the invention, there is provided a kind of data-base cluster Automatic Optimal processing method, it includes：

The status information of monitoring data storehouse cluster, gather the state index of data-base cluster corresponding with preset state index Characteristic information；

According to the state index characteristic information of data-base cluster, the operation conditions of analytical database cluster；And

Corresponding troubleshooting measure and/or optimization processing measure, which are triggered, for operation conditions carries out Automatic Optimal processing.

According to another aspect of the present invention, there is provided a kind of data-base cluster Automatic Optimal processing unit, it includes：

Acquisition module, for the status information of monitoring data storehouse cluster, gather database corresponding with preset state index The state index characteristic information of cluster；

Analysis module, for the state index characteristic information according to data-base cluster, the operation shape of analytical database cluster Condition；And

Processing module, carried out for triggering corresponding troubleshooting measure and/or optimization processing measure for operation conditions Automatic Optimal processing.

According to another aspect of the invention, there is provided a kind of server, including：Processor, memory, communication interface and logical Believe bus, processor, memory and communication interface complete mutual communication by communication bus；

Memory is used to deposit an at least executable instruction, and executable instruction makes the above-mentioned data-base cluster of computing device certainly Operated corresponding to dynamic optimized treatment method.

In accordance with a further aspect of the present invention, there is provided a kind of computer-readable storage medium, be stored with least one in storage medium Executable instruction, executable instruction make computing device be operated as corresponding to above-mentioned data-base cluster Automatic Optimal processing method.

, can according to data-base cluster Automatic Optimal processing method, device, server and the computer-readable storage medium of the present invention It is monitored with the status information to data-base cluster, according to state index characteristic information analysis number corresponding to preset state index According to the operation conditions of storehouse cluster, and trigger corresponding measure for different operation conditions and processing is optimized to data-base cluster Or troubleshooting, to optimize the configuration of resource, compared with prior art, without putting into excessive human resources, reduce manpower Cost, and various failures and problem to be optimized that database occurs can be found in time, these problems are automatically processed, are improved The performance and service quality of database.

Described above is only the general introduction of technical solution of the present invention, in order to better understand the technological means of the present invention, And can be practiced according to the content of specification, and in order to allow above and other objects of the present invention, feature and advantage can Become apparent, below especially exemplified by the embodiment of the present invention.

Brief description of the drawings

By reading the detailed description of hereafter preferred embodiment, it is various other the advantages of and benefit it is common for this area Technical staff will be clear understanding.Accompanying drawing is only used for showing the purpose of preferred embodiment, and is not considered as to the present invention Limitation.And in whole accompanying drawing, identical part is denoted by the same reference numerals.In the accompanying drawings：

Fig. 1 shows the flow signal of data-base cluster Automatic Optimal processing method according to an embodiment of the invention Figure；

Fig. 2 shows the flow signal of data-base cluster Automatic Optimal processing method in accordance with another embodiment of the present invention Figure；

Fig. 3 is shown to be illustrated according to the flow of the data-base cluster Automatic Optimal processing method of another embodiment of the invention Figure；

Fig. 4 is shown to be illustrated according to the flow of the data-base cluster Automatic Optimal processing method of further embodiment of the present invention Figure；

Fig. 5 shows the flow signal of data-base cluster Automatic Optimal processing method according to an embodiment of the invention Figure；

Fig. 6 shows the flow signal of data-base cluster Automatic Optimal processing method in accordance with another embodiment of the present invention Figure；

Fig. 7 is shown to be illustrated according to the flow of the data-base cluster Automatic Optimal processing method of another embodiment of the invention Figure；

Fig. 8 is shown to be illustrated according to the flow of the data-base cluster Automatic Optimal processing method of further embodiment of the present invention Figure；

Fig. 9 shows the flow signal of data-base cluster Automatic Optimal processing method according to an embodiment of the invention Figure；

Figure 10 shows the functional block diagram of data-base cluster Automatic Optimal processing unit according to embodiments of the present invention；

Figure 11 shows a kind of structural representation of server according to embodiments of the present invention.

Embodiment

The exemplary embodiment of the disclosure is more fully described below with reference to accompanying drawings.Although the disclosure is shown in accompanying drawing Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here Limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure Completely it is communicated to those skilled in the art.

Fig. 1 shows the flow signal of data-base cluster Automatic Optimal processing method according to an embodiment of the invention Figure.As shown in figure 1, this method comprises the following steps：

Step S101, the status information of monitoring data storehouse cluster, gather data-base cluster corresponding with preset state index State index characteristic information.

The operation conditions of data-base cluster, including problem to be optimized and/or failure problems, can all be embodied in specific state In index feature information, in order to realize that the Automatic Optimal of data-base cluster is handled, the embodiment of the present invention needs to pre-set State index, by monitoring these state indexs, collects the state index of corresponding data-base cluster as monitored object Characteristic information.Wherein, monitoring is to continue to carry out, to find the problem of data-base cluster is present in time.These default shapes State index includes but are not limited to following index：Access situation, connection number occupancy situation, EMS memory occupation situation, disk size account for With situation, index service condition, network traffics, node state, data query situation and/or node log information.Preset state Failure that the parameter and/or needs that index can optimize as needed are paid close attention to is selected, and the present invention is not especially limited to this.

In the present embodiment, from the status information of the data-base cluster of monitoring, collection is corresponding with preset state index The state index characteristic information of the interior and/or real-time data-base cluster of historical time section, and be stored in database, it is easy to be directed to The running status of the state index characteristic information analysis data-base cluster.

Step S102, according to the state index characteristic information of data-base cluster, the operation conditions of analytical database cluster.

The state index characteristic information of data-base cluster is corresponding with preset state index, and every kind of preset state index is included extremely The state index characteristic information of a few characterization database cluster operation conditions, and corresponding at least one preset state index At least one state index characteristic information can reflect the different performance situation of data-base cluster.

Specifically, state index characteristic information corresponding with access situation includes query rate per second (Queries Per Second, abbreviation QPS) and/or processing number of transactions (Transactions Per Second, abbreviation TPS) per second, wherein, QPS It is to specific inquiry the server how many criterion of handled flow, TPS in one second in data-base cluster Refer to the number for the request that the client computer handled in one second is sent to server, QPS and the QPS more high then representative server of value The data query or the pressure of processing undertaken is bigger.

Specifically, state index characteristic information corresponding with connection number occupancy situation includes the connection of each server node Number, the connection number of each server node represents that the number for accessing the server node is more, corresponding to access pressure just It is bigger, easily cause the situation that can not access the server node.

Specifically, state index characteristic information corresponding with EMS memory occupation situation includes memory usage, memory usage It is higher, it is shown to be that the memory usage of the data-base cluster application is higher, at the same time also indicates that the operating pressure of the internal memory It is bigger.

Specifically, state index characteristic information corresponding with disk size occupancy situation includes disk size occupancy, magnetic The disk size that disk capacity occupancy is higher to be shown to be the data-base cluster application is more not enough.

Specifically, state index characteristic information corresponding with index service condition includes index frequency of use and/or index As a result, wherein, index frequency of use is more high then to represent big to the demand of the index, and indexed results then represent that every index can expire Foot is which query demand, because indexed results corresponding to different indexes there may be overlapping or covering situation.

Specifically, state index characteristic information corresponding with network traffics includes network traffics average value and/or network flow Peak value is measured, network traffics average value and/or network traffics peak value are more high, reflect visit capacity of the client computer to data-base cluster It is bigger.

Specifically, state index characteristic information corresponding with node state be unknown, recovery, down, One kind in rollback, fatal, primary and secondary, wherein, if node state be unknown, One kind in recovery, down and fatal, then show node running status exception.

Specifically, state index characteristic information corresponding with data query situation include inquiry response time and inquire about Number of concurrent, wherein, it is slow inquiry or fast inquiry that the response time length of inquiry, which indicates the inquiry, and during the response of inquiry Between more long then server inquiry velocity it is slower, efficiency is lower；The number of concurrent of inquiry indicates while submits the number of inquiry request Amount, the number of concurrent of inquiry is bigger, then the inquiry pressure of corresponding server is bigger, more easily causes server crash.

Specifically, state index characteristic information corresponding with node log information includes check value match information and/or section Point crash info, if check value match information display matching it is inconsistent, show file data corresponding to node log there occurs Damage；If showing node collapses information in node log information, show that the node is collapsed, it is impossible to externally normal to provide Service.

The state index characteristic information of data-base cluster performance can be reflected by collection, it is possible to achieve to data-base cluster Operation conditions analysis, and then targetedly carry out Automatic Optimal processing.

Step S103, trigger corresponding troubleshooting measure and/or optimization processing measure for operation conditions and carry out automatically Optimization processing.

Data-base cluster inevitably goes wrong during operation and maintenance, in order to ensure data-base cluster Stable service can be externally provided, it is necessary to for it is different the problem of trigger different treatment measures.

Specifically, too high for QPS, TPS, memory usage, disk size occupancy are high, and index utilization rate is low, index As a result repeat, treated present in the data-base cluster of the feature such as network traffics average value and/or network traffics peak value height reflection excellent The triggering optimization processing measure of the problem of change optimizes, so that indexed results repeat as an example, if indexed results repeat, i.e., and at least two The indexed results of an index in index indexed results corresponding to other index cover, then correspond to capped index knot The query function of the index of fruit can be replaced by others index, can be by rope for the pending problem of this repetition index Draw the small index of range of results to delete to discharge more data-base cluster resources, improve the runnability of data-base cluster.

Specifically, mismatched for check value, node collapses, the time of inquiry is long and the number of concurrent of inquiry is more, node Failure problems triggering troubleshooting measure present in the data-base clusters of feature reflection such as abnormal state and/or connection number are excessive Handled, so that check value mismatches as an example, if being shown in Journal node, check value is mismatched, then it is assumed that file is damaged Bad, if recovered not in time to the data in this document, this document can not provide correct data so that can not be external Service is provided, for the failure problems of this file corruption, data recovery measure is taken in time, to ensure that data-base cluster is normal Service is externally provided.

The data-base cluster Automatic Optimal processing method that the present embodiment provides, by entering to the status information of data-base cluster Row monitoring, according to the operation conditions of state index characteristic information analysis data-base cluster corresponding to preset state index, and is directed to Different operation conditions triggers corresponding measure and processing or troubleshooting is optimized to data-base cluster, with prior art phase Than without putting into excessive human resources, reducing human cost, and the various failures that database occurs can be found in time With problem to be optimized, these problems are automatically processed, improve the performance and service quality of database.

Fig. 2 shows the flow signal of data-base cluster Automatic Optimal processing method in accordance with another embodiment of the present invention Figure.This method is used to carry out data recovery for the operation conditions of data-base cluster interior joint abnormal state.As shown in Fig. 2 should Method comprises the following steps：

Step S201, the status information of monitoring data storehouse cluster, the node state of each node in acquisition database cluster Characteristic information.

Specifically, by the status information of monitoring data storehouse cluster, the node state feature for gathering each node in real time is believed Breath.By taking MongoDB data-base clusters as an example, the node in cluster includes host node (also referred to as primary database) and (is also referred to as from node From database), state characteristic information of the host node when normally providing service is primary, and service is normally being provided from node When state characteristic information be secondary, if host node and backup node can not temporarily provide service or be asked in the presence of abnormal Topic, its state characteristic information may be unknown, recovery, down, rollback or fatal.Therefore, node state is special The normal or abnormal operation conditions of reference breath reflection egress.

Step S202, according to node state characteristic information, judge each node in data-base cluster operation conditions whether For exception.

When node can not temporarily provide service or when abnormal problem be present, node state characteristic information be unknown, Recovery, down, rollback or fatal, i.e., the node state that ought be collected be unknown, recovery, down, , it is necessary to which the operation conditions of corresponding node is with the presence or absence of abnormal in analytical database cluster during rollback or fatal.Wherein, Unknown represents that node state is unknown state, and recovery represents that node is in and recovers state, and down represents node not Reachable, rollback represents data just in rollback, at the end of rollback, it will be transferred to recovery or secondary shapes State, fatal represent that error, it is necessary to according to daily record, finds out error reason, does synchronization again.By analyzing above node state spy Reference ceases, if unknown, recovery, down, fatal, then it represents that the operation conditions of node is abnormal.

Step S203, it is abnormal node for operation conditions, suspends the node and service is externally provided；To the number in node According to carrying out recovery processing.

Specifically, it is abnormal node for operation conditions, suspends the node and provide service to service request side, and delete Invalid data in node, a complete data are asked to carry out data recovery process, example to from other nodes or backup node Pulling data carries out data recovery process such as from host node or backup node.

The data-base cluster Automatic Optimal processing method that the present embodiment provides, by each in the cluster of real-time data collection storehouse The node state characteristic information of node, and the operation conditions of node status information decision node accordingly, it is different for operation conditions Normal node carries out recovery processing, and this method can not only find the abnormal node of operation conditions in time, and can be to exception Node targetedly carries out data recovery process, avoids causing the disabled situation of node.

Fig. 3 is shown to be illustrated according to the flow of the data-base cluster Automatic Optimal processing method of another embodiment of the invention Figure.This method is used to carry out troubleshooting for the operation conditions of the slow inquiry in the presence of data-base cluster.As shown in figure 3, This method comprises the following steps：

Step S301, the status information of monitoring data storehouse cluster, the response of data inquiry request in acquisition database cluster Time and number of concurrent.

Slow inquiry is the query type that a kind of consumption resource is more, time-consuming, and with the increasing of the connection number accessed simultaneously More, i.e., number of concurrent increases, and slow inquiry can reduce the treatment effeciency of database.The present embodiment is using data query situation as monitoring pair As the response time of gathered data inquiry request and number of concurrent.

Step S302, according to data query in default response time threshold value and default number of concurrent threshold decision data-base cluster Whether request is pending slow inquiry request.

Specifically, if the response time of data inquiry request exceedes default response time threshold value, the data query is judged Ask as slow inquiry request, and if the number of concurrent of current data inquiry request exceed default number of concurrent threshold value, it is determined that this is slow Inquiry request is pending slow inquiry request.Because slow inquiry request can influence the processing of data inquiry request concurrent therewith Efficiency, cause concurrent data inquiry request to respond card slowly, and if number of concurrent exceedes default number of concurrent threshold value, may lead Data inquiry request is caused to cannot respond to, it is therefore desirable to exceed default response time threshold value, at the same time, number to the response time in time The pending slow inquiry for exceeding default number of concurrent threshold value according to the number of concurrent of inquiry request is handled, please to reduce data query The response time asked, improve the treatment effeciency of database.

Step S303, if data inquiry request is pending slow inquiry request, stop performing the data inquiry request.

The operation for forcing to stop the request being taken for pending slow inquiry request so that concurrent other normal numbers Being capable of normal response according to inquiry request.

The data-base cluster Automatic Optimal processing method that the present embodiment provides, passes through data query in acquisition database cluster The response time of request and number of concurrent, whether it is to inquire about slowly according to default response time threshold decision data inquiry request, and Determine whether slow inquiry request is pending slow according to the number of concurrent of current data inquiry request and default number of concurrent threshold value Inquiry request, the operation for forcing to stop the request being taken for pending slow inquiry request, to avoid because this is pending The influence of slow inquiry request and the problem of cause data inquiry request low-response or even cannot respond to, reduce data inquiry request Response time, improve the treatment effeciency of database.

Fig. 4 is shown to be illustrated according to the flow of the data-base cluster Automatic Optimal processing method of further embodiment of the present invention Figure.This method is used to be directed to file corruption and/or the service data of node collapses in data-base cluster and recovered.As shown in figure 4, should Method comprises the following steps：

Step S401, the status information of monitoring data storehouse cluster, the node log information in acquisition database cluster.

By taking MongoDB data-base clusters as an example, MongoDB is often verified to file to judge the data in file Whether match, and in node log information record matching characteristic information, if for example, data in file mismatch, Unmatched record can be shown in corresponding node log information, according to the record, can just find the feelings of file corruption in time Condition.

Step S402, scan node log information, data file, index text are judged according to the check value of node log information Whether part and/or meta data file damage, and/or, whether collapse case is occurred according to node log information decision node.

In the present embodiment, by scan node log information, and the check value matching in node log information Characteristic information, determines whether file corruption, and file corruption includes data file damage, index file damage and/or metadata File corruption；And/or by scan node log information, and the strace information decision nodes in node log information Whether the collapse case that can not estimate is occurred, wherein, strace information records system calling, it can reflect server The collapse case of node..

Exemplified by judging whether file damages using check value, by compare identical file different times check value come Judge whether this document is damaged, for example, after a certain node is write data to, calculate data in the node file Verification and, next time from the node read data when, calculate again the node file data verification and, if counted twice The verification of calculation and difference, it is determined that the data in file are inconsistent, and the inconsistent characteristic information is recorded in into node log In information, the situation of node file damage can be found in time by scan node log information.

Step S403, if data file, index file and/or meta data file damage, and/or, there are collapse feelings in node Condition, then recovery processing is carried out to the data in node.

If node log information shows data file, index file and/or meta data file damage, and/or, node occurs Collapse case, then need to repair so that this document and/or node can be just the file of damage or the node collapsed Service is often externally provided.

When being repaired, ask a complete data to master library first, then in preset time period to this document and/ Or node operation conditions carry out cycle detection, if monitoring result show it is without exception, then it is assumed that the reparation to this document/or node Complete, if testing result display is abnormal, scans log information again, optionally recovered (equivalent to first by all texts Part empties, then asks a complete data to master library), that is, complete to repair.

The data-base cluster Automatic Optimal processing method that the present embodiment provides, by scan node log information, and is utilized The characteristic information and/or strace information of check value matching result in node log information judge whether that there occurs file corruption And/or node collapses, and file and/or the node of collapse for damaging are repaired, to ensure the section in data-base cluster Point can normally provide service all the time.

Fig. 5 shows the flow signal of data-base cluster Automatic Optimal processing method according to an embodiment of the invention Figure.This method is used to be directed to the operation conditions that connection number is inadequate in data-base cluster and is attached several adjustment.As shown in figure 5, should Method comprises the following steps：

Step S501, the status information of monitoring data storehouse cluster, the node log information in acquisition database cluster.

In data-base cluster in the node log information of each node, the extraneous information for accessing the node is had, wherein Characteristic information including connecting number occupancy situation.

Whether step S502, scan node log information, the connection number of decision node already exceed default connection number threshold value.

Scan node log information, it is current by the characteristic information decision node on connecting number in node log information Connection number whether exceeded default connection number threshold value, for example, having " too many open files " in the node log information Characteristic information, then it represents that exceeded it is default connection number threshold value, subsequently will be due to that can not create for the access request of the node Build connection and cause the failure of no respond request.

Step S503, if the connection number of node alreadys exceed default connection number threshold value, adjust default connection number threshold value.

The situation of default connection number threshold value is alreadyd exceed for connection number, by adjusting default connection number threshold value so that follow-up Access may I ask can normal response, and then the node can not be influenceed service is externally provided.

The data-base cluster Automatic Optimal processing method that the present embodiment provides, passes through the node day in scan database cluster Will information, the characteristic information of the connection number on present node is obtained, it is pre- to judge whether connection number has exceeded for characteristic information accordingly If connection number threshold value, for the situation more than default connection number threshold value, by adjusting default connection number threshold value so that follow-up Access may I ask can normal response, and then the node can not be influenceed service is externally provided.

Fig. 6 shows the flow signal of data-base cluster Automatic Optimal processing method in accordance with another embodiment of the present invention Figure.This method is used to be directed to the inadequate automatic dilatation of operation conditions progress of cluster capacity in data-base cluster.As shown in fig. 6, should Method comprises the following steps：

Step S601, the status information of monitoring data storehouse cluster, network traffics, Operational Visit in acquisition database cluster Feature, I/O utilization rates and/or cluster capacity occupancy situation.

The storage capacity of data-base cluster depends on cluster capacity, and after the operation expanding of service request side, its needs is deposited The data volume of storage can increase therewith, and requirements for access, network traffics and I/O utilization rates can also increase therewith, when cluster capacity not When can meet above-mentioned increased requirement, the situation that read-write efficiency is low or even can not read and write just occurs in data-base cluster, and then can not Normally service is provided to service request side.

In the present embodiment, the state index characteristic information of the data-base cluster on being influenceed by business demand is acquired and profit Whether cluster capacity is rationally judged with the state index characteristic information, the state index characteristic information includes network flow Amount, Operational Visit feature, I/O utilization rates and/or cluster capacity occupancy situation, wherein, Operational Visit feature include TPS and/or QPS。

Step S602, characteristic threshold value, default I/O utilization rate threshold values are accessed according to default network traffics threshold value, pre-set business And/or whether default cluster capacity occupancy situation threshold decision cluster capacity meets current business demand.

Feature threshold is accessed by comparing cell flow and default network traffics threshold value, Operational Visit feature and pre-set business Value, I/O utilization rates and default I/O utilization rates threshold value and/or cluster capacity occupancy situation and default cluster capacity occupancy situation threshold Value, to judge whether cluster capacity meets current business demand.If the network traffics, Operational Visit in data-base cluster are special Have in sign, I/O utilization rates and cluster capacity occupancy situation it is one or more exceed its corresponding predetermined threshold value, then judge cluster Capacity can not meet current business demand.

Step S603, if cluster capacity can not meet current business demand, carry out automatic dilatation.

The situation of current business demand can not be met for cluster capacity, if without dilatation, data-base cluster can not Normally service is provided to service request side.And, it is necessary to network traffics, business in data-base cluster when carrying out automatic dilatation Access feature, I/O utilization rates and/or cluster capacity occupancy situation and exceed the plussage of its corresponding predetermined threshold value, and combine industry The current business demand of requesting party be engaged in determine the amount of dilatation, meets current business with optimal cluster capacity configuration to realize Demand, even if optimal cluster capacity configuration cluster capacity can either meet storage demand and the read-write efficiency of business datum, And can accesses the configuration farthest utilized.

The data-base cluster Automatic Optimal processing method that the present embodiment provides, passes through the network flow in acquisition database cluster Amount, Operational Visit feature, I/O utilization rates and/or cluster capacity occupancy situation；And according to default network traffics threshold value, default industry Whether business accesses characteristic threshold value, default I/O utilization rates threshold value and/or default cluster capacity occupancy situation threshold decision cluster capacity Meet current business demand；The situation of current business demand can not be met for cluster capacity, carry out automatic dilatation, to realize use Optimal cluster capacity configuration meets current business demand.

Fig. 7 is shown to be illustrated according to the flow of the data-base cluster Automatic Optimal processing method of another embodiment of the invention Figure.This method is used to be directed to set meal in data-base cluster and sets irrational operation conditions to carry out set meal level adjustment.Such as Fig. 7 institutes Show, this method comprises the following steps：

Step S701, the status information of monitoring data storehouse cluster, capturing service user account for the capacity of data-base cluster With situation and/or Operational Visit feature.

In database, often apply that during a cluster a corresponding set meal will be matched, set meal is to internal memory, disk valency Lattice, network traffics are taken and the conversion of the progress such as computer room expense, i.e., quantization conversion is carried out to the resource in all set meals, each Individual set meal has one to estimate supporting business amount, for example, handled when DBA has carried out deletion, cleaning of business etc., Or there occurs the variation of machine migration, the resource required for business user can all change, now set meal will be entered Row is adjusted, namely the consumption to business occupancy resource is adjusted.

In the present embodiment, if it is the appearance by the business user of collection to data-base cluster to need to adjust set meal grade Measure occupancy situation and/or Operational Visit feature to judge, likewise, Operational Visit feature includes TPS and/or QPS.

Step S702, according to business user to the capacity occupancy situation of data-base cluster and/or Operational Visit feature Variation tendency, judge whether the set meal rank setting of business user is reasonable.

If business user is to the capacity occupancy situation and/or Operational Visit feature of data-base cluster in preset time period Continuous decrease, then it is assumed that the resource required for business user is in reduced trend, judges that set meal needs to be adjusted to corresponding less The grade of resource；If business user is to the capacity occupancy situation and/or Operational Visit feature of data-base cluster in preset time Persistently rise in section, then it is assumed that the resource required for business user judges that set meal needs to be adjusted to corresponding in the trend increased The grade of more multiple resource.The situation that above two needs to adjust is that set meal rank sets irrational situation.

Step S703, if the set meal rank setting of business user is unreasonable, according to variation tendency to business user Set meal rank be adjusted.

For set meal rank, irrational situation, the capacity occupancy situation according to business user to data-base cluster are set It is that the resource requirement matching of business user is rational and/or the variation tendency of Operational Visit feature correspondingly adjusts set meal grade Set meal.Specifically, the change according to corresponding to the variation tendency of preset time period inner capacities occupancy situation and/or Operational Visit feature Degree, and combine resource situation, the business datum of business user and the read-write requests institute to business datum currently taken Set meal is adjusted to suitable grade by the resource needed, for example, preset time period inner capacities occupancy situation and/or Operational Visit are special Sign is lasting to be risen, and intensity of variation is big, and after occupation condition is considered, set meal grade is adjusted to across multiple grades The grade of corresponding more multiple resource, if intensity of variation is small, the span of corresponding level adjustment is small, may be selected once only to be adjusted to most Neighbouring grade.

The data-base cluster Automatic Optimal processing method that the present embodiment provides, capturing service user is to data-base cluster Capacity occupancy situation and/or Operational Visit feature；The capacity of data-base cluster is accounted for according to business user in preset time period With situation and/or the variation tendency of Operational Visit feature, judge whether the set meal rank setting of business user is reasonable；And it is directed to Irrational situation is set, according to the reading of the resource situation, the business datum and business datum of business user currently taken Set meal is adjusted to suitable grade by the resource required for write request.The present embodiment can be changed in business or other shadows Judge whether the setting of set meal grade reasonable in time in the case of ringing demand resource, and according to the resource situation currently taken, Set meal is adjusted to suitable by the resource required for the business datum of business user and the read-write requests of business datum automatically Grade.

Fig. 8 is shown to be illustrated according to the flow of the data-base cluster Automatic Optimal processing method of further embodiment of the present invention Figure.This method is used to be directed to the irrational operation conditions of parameter setting in data-base cluster and carries out parameter adjustment.As shown in figure 8, This method comprises the following steps：

Step S801, the status information of monitoring data storehouse cluster, gather data-base cluster corresponding with preset state index State index characteristic information.

Each data-base cluster has the parameterized template of a whole set of fixation in initialization, and the parameter of cluster internal is including interior Parameter, disk size parameter and some performance indications are deposited, these parameters or index will be carried out accordingly with the change of business Adjustment optimize the use of resource.

By gathering some data-base cluster state index characteristic informations, for example access situation, EMS memory occupation situation, disk Capacity occupancy situation and network traffics etc., and judge whether the setting of data base set swarm parameter is reasonable using these information.

Step S802, sampling analysis and calculating are carried out to state index characteristic information；Judge that cluster is joined according to result of calculation Whether several settings is reasonable.

Wherein, collecting swarm parameter includes memory parameters, disk size parameter, compress mode parameter, and/or storage engines ginseng Number.By taking memory parameters as an example, by carrying out analysis calculating to internal memory occupancy situation, if EMS memory occupation situation exceedes default internal memory and accounted for With condition threshold, then judge that the setting of memory parameters is unreasonable.

Step S803, if the setting of collection swarm parameter is unreasonable, collection swarm parameter is adjusted according to result of calculation.

For the irrational situation of cluster parameter setting, Automatic Optimal is carried out to collection swarm parameter according to result of calculation.Such as Compress mode in data-base cluster is optimized, acquiescence compress mode is snappy, and this kind of compress mode can meet one As business demand, and CPU usage is relatively low corresponding to which, but compression ratio is not very high, if result of calculation is shown Disk size occupancy situation is high, then it is zlib that can adjust compress mode, the compression ratio highest of the compress mode, can reduce magnetic The occupancy of disk capacity.

The data-base cluster Automatic Optimal processing method that the present embodiment provides, by being adopted to state index characteristic information Sample is analyzed and calculated；Judge whether the setting for collecting swarm parameter is reasonable according to result of calculation；If the setting for collecting swarm parameter is unreasonable, root Collection swarm parameter is adjusted according to result of calculation.The present embodiment can be by monitoring as the change of business is and the state that changes therewith Index feature information, such as situation, EMS memory occupation situation, disk size occupancy situation and network traffics are accessed, cluster is joined Number is adjusted to optimize the use of resource.

Fig. 9 shows the flow signal of data-base cluster Automatic Optimal processing method according to an embodiment of the invention Figure.This method be used for be directed to data-base cluster in repetition index, index utilization rate it is low and/or missing index operation conditions from Dynamic optimization processing.As shown in figure 9, this method comprises the following steps：

Step S901, the status information of monitoring data storehouse cluster, collection index use information.

In MongoDB, if do not indexed, based on each file and therefrom having to scan through in set when reading data The record for meeting querying condition is chosen, the search efficiency of this mode is low-down, especially when handling mass data, is looked into Inquiry will take a long time, and index the efficiency that can greatly then improve inquiry.But if the setting of index is unreasonable, The efficiency of inquiry can still be influenceed and take excessive resource.

Scanning service request side uses the index use information indexed in data-base cluster and the history of service request side Operational Visit situation, the problem of Analytical Index is present.

Step S902, according to index use information and Operational Visit situation, the problem of Analytical Index is present.

According to index use information, analyse whether repetition index be present；And/or according to index use information, Analytical Index Utilization rate；And/or according to Operational Visit situation, analyse whether to need to create new index.

Specifically, in the index use information of existing at least two index, wherein at least one index is corresponding Query Result Query Result corresponding to other index include, disclosure satisfy that bigger inquiry equivalent to other index Demand, in this case, the query function that at least one index is realized can be replaced by other index, if also, in number According to substantial amounts of this repetition index in the cluster of storehouse, be present, then the resource of substantial amounts of data-base cluster, therefore meeting can be taken Influence processing or the response efficiency of cluster；By scanning index use information, it is found that the index utilization rate of some indexes is low, there is one The possible situation of kind is exactly that the low index of these utilization rates is adapted to special function or period, and the current function has been switched off Or do not use, also just embodying corresponding index has the historical record being utilized, but has currently hardly been used, The low index of this index utilization rate can also take the resource of data-base cluster；The result of scanning index use information shows business The frequency of some fields of supplicant access is higher, but does not establish corresponding these and access the index of field, for example, increasing newly After business function, business user is inquired about field corresponding to the function, but also not in time in data-base cluster Middle to establish related index, such case just illustrates that current index is not met by demand of the business user to New function, Lack index.

Step S903, for different problems existing for index, take appropriate measures.

For repetition index, delete processing is carried out；And/or delete the rope that index utilization rate is less than default utilization rate threshold value Draw；And/or be higher than the access field of default accesss degree threshold value for Operational Visit degree, create it is with access field corresponding newly Index.

Specifically, in order to avoid repetition index and/or occupancy of the index to the resource of data-base cluster of utilization rate is indexed, Delete and meet that the needs of business user small repetition index and/or utilization rate are less than the index of default utilization rate threshold value；Also, In order to meet the new requirements for access of business user, such as the access to some fields, for Operational Visit degree higher than default The access field of access degree threshold value, creates new index.

The data-base cluster Automatic Optimal processing method that the present embodiment provides, by scanning index use information, according to rope Draw use information, analyse whether repetition index be present, and delete the repetition index for meeting that the needs of business user is small；According to rope Draw use information, Analytical Index utilization rate, delete the index that utilization rate is less than default utilization rate threshold value；According to Operational Visit feelings Condition, analyse whether to need to create new index, create Operational Visit degree and be higher than corresponding to the access field of default access degree threshold value Index.The processing mode of above-mentioned deletion indexes occupancy data-base cluster resource corresponding to can reducing, and improves data-base cluster Processing or response efficiency, the mode that above-mentioned establishment newly indexes disclosure satisfy that the new requirements for access of business user.

Figure 10 shows the functional block diagram of data-base cluster Automatic Optimal processing unit according to embodiments of the present invention.Such as figure Shown in 10, the device includes：Acquisition module 11, analysis module 12 and processing module 13.

Acquisition module 11, suitable for the status information of monitoring data storehouse cluster, gather data corresponding with preset state index The state index characteristic information of storehouse cluster.

Wherein, preset state index includes：Access situation, connection number occupancy situation, EMS memory occupation situation, disk size account for With situation, index service condition, network traffics, node state, data query situation and/or node log information.

Acquisition module 11 is further adapted for：In collection historical time section corresponding with preset state index and/or in real time The state index characteristic information of data-base cluster.

Analysis module 12, suitable for the state index characteristic information according to data-base cluster, the operation of analytical database cluster Situation；

Processing module 13, enter suitable for triggering corresponding troubleshooting measure and/or optimization processing measure for operation conditions The processing of row Automatic Optimal.

Operation conditions for data-base cluster interior joint abnormal state in the specific embodiment of the present invention is entered During row data recovery：

Acquisition module 11 is further adapted for：The node state characteristic information of each node in acquisition database cluster；

Analysis module 12 is further adapted for：According to node state characteristic information, each node in data-base cluster is judged Whether operation conditions is abnormal.

Processing module 13 is further adapted for：It is abnormal node for operation conditions, suspends the node and service is externally provided； Recovery processing is carried out to the data in node.

The operation conditions inquired about slowly in data-base cluster is directed in the specific embodiment of the present invention and carries out failure During processing：

Acquisition module 11 is further adapted for：The response time of data inquiry request and number of concurrent in acquisition database cluster；

Analysis module 12 is further adapted for：According to default response time threshold value and default number of concurrent threshold decision data base set Whether data inquiry request is pending slow inquiry request in group.

Processing module 13 is further adapted for：If data inquiry request is pending slow inquiry request, stop performing the number According to inquiry request.

File corruption and/or the fortune of node collapses in data-base cluster are directed in the specific embodiment of the present invention During row data recovery：

Acquisition module 11 is further adapted for：Node log information in acquisition database cluster；

Analysis module 12 is further adapted for：Scan node log information, number is judged according to the check value of node log information Whether damaged according to file, index file and/or meta data file, and/or, whether occurred according to node log information decision node Collapse case.

Processing module 13 is further adapted for：If data file, index file and/or meta data file damage, and/or, section There is collapse case in point, then carries out recovery processing to the data in node.

It is directed in the specific embodiment of the present invention and the inadequate operation conditions progress of number is connected in data-base cluster When connecting number adjustment：

Analysis module 12 is further adapted for：Scan node log information, it is pre- whether the connection number of decision node alreadys exceed If connect number threshold value.

Processing module 13 is further adapted for：If the connection number of node alreadys exceed default connection number threshold value, adjustment is default Connect number threshold value.

The operation conditions that cluster capacity is inadequate in data-base cluster is directed in the specific embodiment of the present invention to enter During the automatic dilatation of row：

Acquisition module 11 is further adapted for：Network traffics, Operational Visit feature, I/O in acquisition database cluster use Rate and/or cluster capacity occupancy situation；

Analysis module 12 is further adapted for：Characteristic threshold value, default I/ are accessed according to default network traffics threshold value, pre-set business Whether O utilization rates threshold value and/or default cluster capacity occupancy situation threshold decision cluster capacity meet current business demand；

Processing module 13 is further adapted for：If cluster capacity can not meet current business demand, automatic dilatation is carried out.

Set meal in data-base cluster is directed in the specific embodiment of the present invention irrational operation conditions is set When carrying out set meal level adjustment：

Acquisition module 11 is further adapted for：Capacity occupancy situation and/or industry of the capturing service user to data-base cluster Business accesses feature；

Analysis module 12 is further adapted for：Capacity occupancy situation and/or industry according to business user to data-base cluster Business accesses the variation tendency of feature, judges whether the set meal rank setting of business user is reasonable.

Processing module 13 is further adapted for：If the set meal rank setting of business user is unreasonable, according to variation tendency The set meal rank of business user is adjusted.

The irrational operation conditions of parameter setting in data-base cluster is directed in the specific embodiment of the present invention When carrying out parameter adjustment：

Analysis module 12 is further adapted for：Sampling analysis and calculating are carried out to state index characteristic information；Tied according to calculating Fruit judges whether the setting for collecting swarm parameter is reasonable, wherein, collection swarm parameter includes memory parameters, disk size parameter, compress mode Parameter, and/or storage engines parameter.

Processing module 13 is further adapted for：If the setting for collecting swarm parameter is unreasonable, collection swarm parameter is adjusted according to result of calculation.

Repetition index, the index utilization rate being directed in the specific embodiment of the present invention in data-base cluster are low And/or when lacking the operation conditions Automatic Optimal processing of index：

Acquisition module 11 is further adapted for：Collection index use information；

Analysis module 12 is further adapted for：According to index use information, analyse whether repetition index be present；And/or according to Index use information, Analytical Index utilization rate；And/or according to Operational Visit situation, analyse whether to need to create new index.

Processing module 13 is further adapted for：For repetition index, delete processing is carried out；And/or deletion index utilization rate is low In the index of default utilization rate threshold value；And/or it is higher than the access field of default access degree threshold value for Operational Visit degree, create New index corresponding with accessing field.

The embodiment of the present application provides a kind of nonvolatile computer storage media, and computer-readable storage medium is stored with least One executable instruction, the computer executable instructions can perform the data-base cluster Automatic Optimal in above-mentioned any means embodiment Processing method.

Figure 11 shows a kind of structural representation of server according to embodiments of the present invention, and the specific embodiment of the invention is simultaneously The specific implementation to server does not limit.

As shown in figure 11, the server can include：Processor (processor) 111, communication interface (Communications Interface) 113, memory (memory) 115 and communication bus 117.

Wherein：

Processor 111, communication interface 113 and memory 115 complete mutual communication by communication bus 117.

Communication interface 113, for being communicated with the network element of miscellaneous equipment such as client or other servers etc..

Processor 111, for configuration processor 119, it can specifically perform above-mentioned data-base cluster Automatic Optimal processing method Correlation step in embodiment.

Specifically, program 119 can include program code, and the program code includes computer-managed instruction.

Processor 111 is probably central processor CPU, or specific integrated circuit ASIC (Application Specific Integrated Circuit), or it is arranged to implement the integrated electricity of one or more of the embodiment of the present invention Road.The one or more processors that server includes, can be same type of processor, such as one or more CPU；Can also It is different types of processor, such as one or more CPU and one or more ASIC.

Memory 115, for depositing program 119.Memory 115 may include high-speed RAM memory, it is also possible to also include Nonvolatile memory (non-volatile memory), for example, at least a magnetic disk storage.

Program 119 specifically can be used for so that processor 111 performs following operation：

According to the state index characteristic information of data-base cluster, the operation conditions of analytical database cluster；

Program 119 specifically can be also used for so that processor 111 performs following operation：

The state index of the interior and/or real-time data-base cluster of collection historical time section corresponding with preset state index is special Reference ceases.

The node state characteristic information of each node in acquisition database cluster；

According to node state characteristic information, whether the operation conditions for judging each node in data-base cluster is abnormal.

It is abnormal node for operation conditions, suspends the node and service is externally provided；

Recovery processing is carried out to the data in node.

The response time of data inquiry request and number of concurrent in acquisition database cluster；

According to data inquiry request in default response time threshold value and default number of concurrent threshold decision data-base cluster whether For pending slow inquiry request.

If data inquiry request is pending slow inquiry request, stop performing the data inquiry request.

Node log information in acquisition database cluster；

Scan node log information, data file, index file and/or member are judged according to the check value of node log information Whether data file is damaged, and/or, whether collapse case is occurred according to node log information decision node.

If data file, index file and/or meta data file damage, and/or, there is collapse case in node, then to section Data in point carry out recovery processing.

Node log information in acquisition database cluster；

Whether scan node log information, the connection number of decision node already exceed default connection number threshold value.

If the connection number of node alreadys exceed default connection number threshold value, default connection number threshold value is adjusted.

Network traffics, Operational Visit feature, I/O utilization rates and/or cluster capacity in acquisition database cluster take feelings Condition；

Characteristic threshold value, default I/O utilization rates threshold value are accessed according to default network traffics threshold value, pre-set business and/or preset Whether cluster capacity occupancy situation threshold decision cluster capacity meets current business demand.

If cluster capacity can not meet current business demand, automatic dilatation is carried out.

Capacity occupancy situation and/or Operational Visit feature of the capturing service user to data-base cluster；

According to business user to the capacity occupancy situation of data-base cluster and/or the variation tendency of Operational Visit feature, Judge whether the set meal rank setting of business user is reasonable.

If the set meal rank of business user sets unreasonable, the set meal rank according to variation tendency to business user It is adjusted.

Sampling analysis and calculating are carried out to state index characteristic information；

Judge whether the setting for collecting swarm parameter is reasonable according to result of calculation, wherein, collection swarm parameter includes memory parameters, disk Capacity parameter, compress mode parameter, and/or storage engines parameter.

If the setting for collecting swarm parameter is unreasonable, collection swarm parameter is adjusted according to result of calculation.

Collection index use information；

According to index use information, analyse whether repetition index be present；

And/or according to index use information, Analytical Index utilization rate；

And/or according to Operational Visit situation, analyse whether to need to create new index.

For repetition index, delete processing is carried out；

And/or delete the index that index utilization rate is less than default utilization rate threshold value；

And/or it is higher than the access field of default access degree threshold value for Operational Visit degree, create corresponding with accessing field New index.

Algorithm and display be not inherently related to any certain computer, virtual system or miscellaneous equipment provided herein. Various general-purpose systems can also be used together with teaching based on this.As described above, required by constructing this kind of system Structure be obvious.In addition, the present invention is not also directed to any certain programmed language.It should be understood that it can utilize various Programming language realizes the content of invention described herein, and the description done above to language-specific is to disclose this hair Bright preferred forms.

In the specification that this place provides, numerous specific details are set forth.It is to be appreciated, however, that the implementation of the present invention Example can be put into practice in the case of these no details.In some instances, known method, structure is not been shown in detail And technology, so as not to obscure the understanding of this description.

Similarly, it will be appreciated that in order to simplify the disclosure and help to understand one or more of each inventive aspect, Above in the description to the exemplary embodiment of the present invention, each feature of the invention is grouped together into single implementation sometimes In example, figure or descriptions thereof.However, the method for the disclosure should be construed to reflect following intention：I.e. required guarantor The application claims of shield features more more than the feature being expressly recited in each claim.It is more precisely, such as following Claims reflect as, inventive aspect is all features less than single embodiment disclosed above.Therefore, Thus the claims for following embodiment are expressly incorporated in the embodiment, wherein each claim is in itself Separate embodiments all as the present invention.

Those skilled in the art, which are appreciated that, to be carried out adaptively to the module in the equipment in embodiment Change and they are arranged in one or more equipment different from the embodiment.Can be the module or list in embodiment Member or component be combined into a module or unit or component, and can be divided into addition multiple submodule or subelement or Sub-component.In addition at least some in such feature and/or process or unit exclude each other, it can use any Combination is disclosed to all features disclosed in this specification (including adjoint claim, summary and accompanying drawing) and so to appoint Where all processes or unit of method or equipment are combined.Unless expressly stated otherwise, this specification (including adjoint power Profit requires, summary and accompanying drawing) disclosed in each feature can be by providing the alternative features of identical, equivalent or similar purpose come generation Replace.

In addition, it will be appreciated by those of skill in the art that although some embodiments in this include institute in other embodiments Including some features rather than further feature, but the combination of the feature of different embodiments means to be in the scope of the present invention Within and form different embodiments.For example, in the following claims, embodiment claimed it is any it One mode can use in any combination.

The all parts embodiment of the present invention can be realized with hardware, or to be run on one or more processor Software module realize, or realized with combinations thereof.It will be understood by those of skill in the art that it can use in practice Microprocessor or digital signal processor (DSP) realize data-base cluster Automatic Optimal processing according to embodiments of the present invention The some or all functions of some or all parts in device.The present invention is also implemented as being used to perform being retouched here The some or all equipment or program of device (for example, computer program and computer program product) for the method stated. Such program for realizing the present invention can store on a computer-readable medium, or can have one or more signal Form.Such signal can be downloaded from internet website and obtained, either provide on carrier signal or with it is any its He provides form.

It should be noted that the present invention will be described rather than limits the invention for above-described embodiment, and ability Field technique personnel can design alternative embodiment without departing from the scope of the appended claims.In the claims, Any reference symbol between bracket should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not Element or step listed in the claims.Word "a" or "an" before element does not exclude the presence of multiple such Element.The present invention can be by means of including the hardware of some different elements and being come by means of properly programmed computer real It is existing.In if the unit claim of equipment for drying is listed, several in these devices can be by same hardware branch To embody.The use of word first, second, and third does not indicate that any order.These words can be explained and run after fame Claim.

The invention discloses a kind of data-base cluster Automatic Optimal processing methods of A1., including：

According to the state index characteristic information of the data-base cluster, the operation conditions of analytical database cluster；And

A2. the method according to A1, wherein, the preset state index includes：Access situation, connection number take feelings Condition, EMS memory occupation situation, disk size occupancy situation, index service condition, network traffics, node state, data query situation And/or node log information.

A3. the method according to A1 or A2, wherein, corresponding with the preset state index data-base cluster of the collection State index characteristic information further comprises：Collection historical time section corresponding with preset state index is interior and/or counts in real time According to the state index characteristic information of storehouse cluster.

A4. the method according to any one of A1-A3, wherein, it is described to gather database corresponding with preset state index The state index characteristic information of cluster further comprises：The node state characteristic information of each node in acquisition database cluster；

The state index characteristic information according to data-base cluster, the operation conditions of analytical database cluster are further wrapped Include：According to node state characteristic information, whether the operation conditions for judging each node in the data-base cluster is abnormal.

A5. the method according to A4, wherein, it is described to trigger corresponding troubleshooting measure progress certainly for operation conditions Dynamic optimization processing further comprises：

It is abnormal node for operation conditions, suspends the node and service is externally provided；And

Recovery processing is carried out to the data in the node.

A6. the method according to any one of A1-A3, wherein, it is described to gather database corresponding with preset state index The state index characteristic information of cluster further comprises：Response time of data inquiry request and concurrent in acquisition database cluster Number；

The state index characteristic information according to data-base cluster, the operation conditions of analytical database cluster are further wrapped Include：According to data inquiry request in data-base cluster described in default response time threshold value and default number of concurrent threshold decision whether be Pending slow inquiry request.

A7. the method according to A6, wherein, it is described to trigger corresponding troubleshooting measure progress certainly for operation conditions Dynamic optimization processing further comprises：If data inquiry request is pending slow inquiry request, stopping performs the data query please Ask.

A8. the method according to any one of A1-A3, wherein, it is described to gather database corresponding with preset state index The state index characteristic information of cluster further comprises：Node log information in acquisition database cluster；

The state index characteristic information according to data-base cluster, the operation conditions of analytical database cluster are further wrapped Include：

A9. the method according to A8, wherein, it is described to trigger corresponding troubleshooting measure progress certainly for operation conditions Dynamic optimization processing further comprises：

If the data file, index file and/or meta data file damage, and/or, there is collapse case in node, then Recovery processing is carried out to the data in the node.

A10. the method according to any one of A1-A3, wherein, it is described to gather database corresponding with preset state index The state index characteristic information of cluster further comprises：Node log information in acquisition database cluster；

The state index characteristic information according to data-base cluster, the operation conditions of analytical database cluster are further wrapped Include：Whether scan node log information, the connection number of decision node already exceed default connection number threshold value.

A11. the method according to A10, wherein, it is described to trigger corresponding troubleshooting measure progress for operation conditions Automatic Optimal processing further comprises：If the connection number of node alreadys exceed default connection number threshold value, the default company is adjusted Connect several threshold values.

A12. the method according to any one of A1-A3, wherein, it is described to gather database corresponding with preset state index The state index characteristic information of cluster further comprises：Network traffics, Operational Visit feature in acquisition database cluster, I/O Utilization rate and/or cluster capacity occupancy situation；

The state index characteristic information according to data-base cluster, the operation conditions of analytical database cluster are further wrapped Include：Characteristic threshold value, default I/O utilization rates threshold value and/or default cluster are accessed according to default network traffics threshold value, pre-set business to hold Whether amount occupancy situation threshold decision cluster capacity meets current business demand.

A13. the method according to A12, wherein, it is described to trigger corresponding optimization processing measure progress for operation conditions Automatic Optimal processing further comprises：If cluster capacity can not meet current business demand, automatic dilatation is carried out.

A14. the method according to any one of A1-A3, wherein, it is described to gather data corresponding with preset state index The state index characteristic information of storehouse cluster further comprises：Capacity occupancy situation of the capturing service user to data-base cluster And/or Operational Visit feature；

The state index characteristic information according to data-base cluster, the operation conditions of analytical database cluster are further wrapped Include：According to business user to the capacity occupancy situation of data-base cluster and/or the variation tendency of Operational Visit feature, institute is judged Whether the set meal rank setting for stating business user is reasonable.

A15. the method according to A14, wherein, it is described to trigger corresponding optimization processing measure progress for operation conditions Automatic Optimal processing further comprises：If the set meal rank setting of the business user is unreasonable, become according to the change Gesture is adjusted to the set meal rank of the business user.

A16. the method according to any one of A1-A3, wherein, it is described to be believed according to the state index feature of data-base cluster Breath, the operation conditions of analytical database cluster further comprise：

Sampling analysis and calculating are carried out to the state index characteristic information；And

A17. the method according to A16, wherein, it is described automatic excellent for operation conditions triggering optimization processing measure progress Change processing further comprises：

A18. the method according to any one of A1-3, wherein, it is described to gather database corresponding with preset state index The state index characteristic information of cluster further comprises：Collection index use information；

The state index characteristic information according to the data-base cluster, the operation conditions of analytical database cluster enter one Step includes：

And/or according to index use information, Analytical Index utilization rate；

A19. the method according to A18, wherein, it is described to trigger corresponding optimization processing measure progress for operation conditions Automatic Optimal processing further comprises：

For repetition index, delete processing is carried out；

And/or it is higher than the access field of default access degree threshold value for Operational Visit degree, create and the access field pair The new index answered.

The invention also discloses a kind of data-base cluster Automatic Optimal processing units of B20., including：

Acquisition module, suitable for the status information of monitoring data storehouse cluster, gather database corresponding with preset state index The state index characteristic information of cluster；

Analysis module, suitable for the state index characteristic information according to the data-base cluster, the fortune of analytical database cluster Row situation；And

Processing module, carried out suitable for triggering corresponding troubleshooting measure and/or optimization processing measure for operation conditions Automatic Optimal processing.

B21. the device according to B20, wherein, the preset state index includes：Access situation, connection number take feelings Condition, EMS memory occupation situation, disk size occupancy situation, index service condition, network traffics, node state, data query situation And/or node log information.

B22. the device according to B20 or B21, wherein, the acquisition module is further adapted for：Collection and preset state The state index characteristic information of the interior and/or real-time data-base cluster of historical time section corresponding to index.

B23. the device according to any one of B20-B22, wherein, the acquisition module is further adapted for：Gathered data The node state characteristic information of each node in the cluster of storehouse；

The analysis module is further adapted for：According to node state characteristic information, judge each in the data-base cluster Whether the operation conditions of node is abnormal.

B24. the device according to B23, wherein, the processing module is further adapted for：

Recovery processing is carried out to the data in the node.

B25. the device according to any one of B20-B22, wherein, the acquisition module is further adapted for：Gathered data The response time of data inquiry request and number of concurrent in the cluster of storehouse；

The analysis module is further adapted for：The number according to default response time threshold value and default number of concurrent threshold decision Whether it is pending slow inquiry request according to data inquiry request in the cluster of storehouse.

B26. the device according to B25, wherein, the processing module is further adapted for：If data inquiry request is to treat The slow inquiry request of processing, stop performing the data inquiry request.

B27. the device according to any one of B20-B22, wherein, the acquisition module is further adapted for：Gathered data Node log information in the cluster of storehouse；

The analysis module is further adapted for：Scan node log information, judged according to the check value of node log information Whether data file, index file and/or meta data file damage, and/or, whether gone out according to node log information decision node Existing collapse case.

B28. the device according to B27, wherein, the processing module is further adapted for：

If data file, index file and/or meta data file damage, and/or, there is collapse case in node, then to institute The data stated in node carry out recovery processing.

B29. the device according to any one of B20-B22, wherein, the acquisition module is further adapted for：Gathered data Node log information in the cluster of storehouse；

The analysis module is further adapted for：Whether scan node log information, the connection number of decision node already exceed Default connection number threshold value.

B30. the device according to B29, wherein, the processing module is further adapted for：If the connection number of node is More than default connection number threshold value, then the default connection number threshold value is adjusted.

B31. the device according to any one of B20-B22, wherein, the acquisition module is further adapted for：Gathered data Network traffics, Operational Visit feature, I/O utilization rates and/or cluster capacity occupancy situation in the cluster of storehouse；

The analysis module is further adapted for：Characteristic threshold value is accessed according to default network traffics threshold value, pre-set business, preset Whether I/O utilization rates threshold value and/or default cluster capacity occupancy situation threshold decision cluster capacity meet current business demand；

B32. the device according to B31, wherein, the processing module is further adapted for：If cluster capacity can not meet Current business demand, then carry out automatic dilatation.

B33. the device according to any one of B20-B22, wherein, the acquisition module is further adapted for：Gathering Capacity occupancy situation and/or Operational Visit feature of the business user to data-base cluster；

The analysis module is further adapted for：According to business user to the capacity occupancy situation of data-base cluster and/or The variation tendency of Operational Visit feature, judge whether the set meal rank setting of the business user is reasonable.

B34. the device according to B33, wherein, the processing module is further adapted for：If the business user's The setting of set meal rank is unreasonable, then the set meal rank of the business user is adjusted according to the variation tendency.

B35. the device according to any one of B20-B22, wherein, the analysis module is further adapted for：

B36. the device according to B35, wherein, the processing module is further adapted for：

B37. the device according to any one of B20-B22, wherein, the acquisition module is further adapted for：Collection index Use information；

The analysis module is further adapted for：

And/or according to index use information, Analytical Index utilization rate；

B38. the device according to B37, wherein, the processing module is further adapted for：

For repetition index, delete processing is carried out；

The invention also discloses a kind of servers of C39., including：Processor, memory, communication interface and communication bus, institute State processor, the memory and the communication interface and mutual communication is completed by the communication bus；

The memory is used to deposit an at least executable instruction, and the executable instruction makes the computing device such as Operated corresponding to data-base cluster Automatic Optimal processing method any one of A1-A19.

The invention also discloses a kind of computer-readable storage mediums of D40., it is executable that at least one is stored with the storage medium Instruction, the executable instruction make the data-base cluster Automatic Optimal processing method any one of computing device A1-A19 Corresponding operation.

Claims

1. a kind of data-base cluster Automatic Optimal processing method, including：

The status information of monitoring data storehouse cluster, gather the state index feature of data-base cluster corresponding with preset state index Information；

2. according to the method for claim 1, wherein, the preset state index includes：Access situation, connection number take feelings Condition, EMS memory occupation situation, disk size occupancy situation, index service condition, network traffics, node state, data query situation And/or node log information.

3. method according to claim 1 or 2, wherein, it is described to gather data-base cluster corresponding with preset state index State index characteristic information further comprise：In collection historical time section corresponding with preset state index and/or in real time The state index characteristic information of data-base cluster.

4. according to the method described in claim any one of 1-3, wherein, it is described to gather database corresponding with preset state index The state index characteristic information of cluster further comprises：The node state characteristic information of each node in acquisition database cluster；

The state index characteristic information according to data-base cluster, the operation conditions of analytical database cluster further comprise： According to node state characteristic information, whether the operation conditions for judging each node in the data-base cluster is abnormal.

5. according to the method for claim 4, wherein, the operation conditions that is directed to triggers corresponding troubleshooting measure progress Automatic Optimal processing further comprises：

Recovery processing is carried out to the data in the node.

6. according to the method described in claim any one of 1-3, wherein, it is described to gather database corresponding with preset state index The state index characteristic information of cluster further comprises：Response time of data inquiry request and concurrent in acquisition database cluster Number；

The state index characteristic information according to data-base cluster, the operation conditions of analytical database cluster further comprise： Whether it is to treat according to data inquiry request in data-base cluster described in default response time threshold value and default number of concurrent threshold decision The slow inquiry request of processing.

7. according to the method for claim 6, wherein, the operation conditions that is directed to triggers corresponding troubleshooting measure progress Automatic Optimal processing further comprises：If data inquiry request is pending slow inquiry request, stop performing the data query Request.

8. a kind of data-base cluster Automatic Optimal processing unit, including：

Acquisition module, suitable for the status information of monitoring data storehouse cluster, gather data-base cluster corresponding with preset state index State index characteristic information；

Analysis module, suitable for the state index characteristic information according to the data-base cluster, the operation shape of analytical database cluster Condition；And

Processing module, carried out automatically suitable for triggering corresponding troubleshooting measure and/or optimization processing measure for operation conditions Optimization processing.

9. a kind of server, including：Processor, memory, communication interface and communication bus, the processor, the memory Mutual communication is completed by the communication bus with the communication interface；

The memory is used to deposit an at least executable instruction, and the executable instruction makes the computing device such as right will Ask and operated corresponding to the data-base cluster Automatic Optimal processing method any one of 1-7.

10. a kind of computer-readable storage medium, an at least executable instruction, the executable instruction are stored with the storage medium Make operation corresponding to data-base cluster Automatic Optimal processing method of the computing device as any one of claim 1-7.