CN108366094A - A kind of cloud storage platform management method - Google Patents
A kind of cloud storage platform management method Download PDFInfo
- Publication number
- CN108366094A CN108366094A CN201810041455.6A CN201810041455A CN108366094A CN 108366094 A CN108366094 A CN 108366094A CN 201810041455 A CN201810041455 A CN 201810041455A CN 108366094 A CN108366094 A CN 108366094A
- Authority
- CN
- China
- Prior art keywords
- storage device
- reliability
- cluster
- user
- region
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1097—Protocols in which an application is distributed across nodes in the network for distributed storage of data in networks, e.g. transport arrangements for network file system [NFS], storage area networks [SAN] or network attached storage [NAS]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/101—Server selection for load balancing based on network conditions
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1012—Server selection for load balancing based on compliance of requirements or conditions with available server resources
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention provides a kind of cloud storage platform management method, including：S1, if the user storage device that cloud storage cluster is added in application meets preset hardware requirement, then the user storage device is added in the storage device cluster of IP ownership place corresponding region of the user storage device, and according on the capacity of the user storage device and reliability based distribution corresponding PG to the user storage device.A kind of cloud storage platform management method provided by the invention, cloud storage platform is added in user storage device by the IP, capacity and reliability of user storage device, solve the problems, such as the idle of user's local resource, the utilization rate of idle disk is improved, while space cost is saved for cloud storage platform.
Description
Technical field
The present invention relates to field of cloud computer technology more particularly to a kind of cloud storage platform management methods.
Background technology
Cloud storage be in the conceptive extension of cloud computing (cloud computing) and a new concept developed,
It is a kind of emerging Network storage technology, refers to by functions such as cluster application, network technology or distributed file systems, by net
A large amount of various types of storage devices gather collaborative work by application software in network, and the common data that externally provide are deposited
The system of storage and Operational Visit function.In simple terms, cloud storage is exactly that storage resources are put into the one kind accessed for people on cloud is new
Emerging scheme.User can at any time, Anywhere be connected to through any web-enabled device on cloud and easily access
Data.
Increasingly mature with cloud computing technology, the advantage of cloud storage is increasingly notable, and family's disk is in terms of convenience
The drawbacks of also gradually highlight.Lead to the damage of the loss and disk of data without backup consciousness and without special protection mechanism memory,
And thus cloud storage also can will lead to people couple by it with the only choosing for becoming user with the convenience and reliability taken is deposited
The demand diminuendo of traditional-family's disk, utilization rate are also lower and lower, and the demand to cloud storage is increasing, this commonly to use
The idle disk space of family family is increasing, causes a large amount of wasting of resources.
Invention content
The present invention is to solve the problems, such as that the idle disk of family existing in the prior art leads to the wasting of resources, provides one kind
Cloud storage platform management method.
The present invention proposes a kind of cloud storage platform management method, including：S1, if the user that cloud storage cluster is added in application deposits
Storage equipment meets preset hardware requirement, then the IP that the user storage device is added to the user storage device belongs to
In the storage device cluster of ground corresponding region, and arrived according to the capacity of the user storage device and the corresponding PG of reliability based distribution
In the user storage device.
Preferably, further include：S0 builds the memory device set of the IP ownership place corresponding region of the user storage device
Group；The step S0 further comprises：S01, if the level where IP ownership place corresponding region further includes other regions, and
The load of the storage device cluster in any other regions meets preset loading condition, then from other described any areas
Any user storage device is exported in the storage device cluster in domain as remote backup equipment；Otherwise, from main storage device cluster
Middle output any server storage device is as remote backup equipment；The main storage device cluster is set by multiple server storages
Standby to constitute, the server storage device is the storage device that service provider provides；S02, according to the capacity of each user storage device
And reliability, choose several user storage devices from the user storage device under IP ownership place corresponding region；S03,
The sum of the reliability of the remote backup equipment and several user storage devices is calculated, if the sum of described reliability is not less than pre-
The reliability cluster threshold value first set then builds the IP using the remote backup equipment and several user storage devices and returns
The storage device cluster of possession corresponding region.
Preferably, further include：S2, according to the reliability of each storage device, network in the storage device cluster in the region
At least one of delay and disk performance, choose the main storage device of the storage device cluster in the region.
Preferably, further include：If in the storage device cluster in the region, the reliability of any user storage device is less than
Preset minimum reliability threshold value then deletes any user storage device from the storage device cluster.
Preferably, the step S03 further includes：If the sum of described reliability is less than preset reliability cluster threshold
It is worth, then the export server storage device from the main storage device cluster, so that after the server storage device is added
The sum of reliability be not less than preset reliability cluster threshold value, set using the server storage device, remote backup
The storage device cluster of standby and described several storage devices structure current region.
Preferably, a reliability is included at least in the storage device cluster be not less than preset reliable apparatus threshold
The storage device of value；The reliability of the server storage device immobilizes and is more than the reliable apparatus threshold value.
Preferably, described if the user storage device that cloud storage cluster is added in application meets preset hardware requirement
Step S1 further includes：To the user storage device carry out sequence read-write and random read-write, the user storage device is obtained
Read or write speed and IOPS values.
Preferably, further include：S3 after waiting for preset duration, updates in the storage device cluster in the region respectively storage and sets
Standby reliability executes step S2 at least once.
Preferably, when the reliability of each storage device, network delay and magnetic in the storage device cluster according to the region
Disk performance, when choosing the main storage device of storage device cluster in the region, the step S2 further comprises：S21, extraction
Reliability is not less than each storage device of preset reliable apparatus threshold value in the storage device cluster in the region；S22 is answered
The network delay of each storage device described in form calculus with random sampling point ping packets is with lowest latency in the network delay
Benchmark converts the corresponding network delay score for obtaining each storage device by percentage；With the IOPS values of each storage device
Middle highest IOPS values convert the corresponding disk performance score for obtaining each storage device as benchmark, by percentage；S23, by institute
Comprehensive score of the sum of network delay score and the disk performance score of each storage device as each storage device is stated, described in extraction
The highest storage device of comprehensive score is the main storage device of the storage device cluster in the region in each storage device.
Preferably, the reliability of the user storage device is calculated by following formula and is obtained：
Reliability=a reference value × disk life value × availability × length of delay+disk performance
Wherein, the disk life value ∈ [0,1], the availability ∈ [0,1], the length of delay ∈ [0,1].
A kind of cloud storage platform management method provided by the invention, passes through the IP, capacity and reliability of user storage device
Cloud storage platform is added in user storage device, the idle of user's local resource is solved the problems, such as, improves the profit of idle disk
Space cost is saved with rate, while for cloud storage platform.
Description of the drawings
Fig. 1 is a kind of structural schematic diagram of cloud storage platform of the specific embodiment of the invention；
Fig. 2 is a kind of flow diagram of cloud storage platform management method of the specific embodiment of the invention.
Specific implementation mode
With reference to the accompanying drawings and examples, the specific implementation mode of the present invention is described in further detail.Implement below
Example is not limited to the scope of the present invention for illustrating the present invention.
A kind of cloud storage platform management method, including：S1, if the user storage device that cloud storage cluster is added in application meets
The IP ownership place corresponding region of the user storage device is then added in the user storage device by preset hardware requirement
Storage device cluster in, and deposited to the user according to the capacity of the user storage device and the corresponding PG of reliability based distribution
It stores up in equipment.
Specifically, the cloud storage platform is established based on Ceph distributed file systems.In the specific embodiment of the invention, institute
It states in addition to server storage device in cloud storage platform, further includes user storage device, described in the user storage device addition
The method of cloud storage platform is as follows：
First, it receives application and the request that the user storage device of cloud storage cluster is sent is added, and judge that the user deposits
Whether storage equipment meets preset hardware requirement：
For example, pre-set hardware standard is 100GB+1 core+200MB memories.The multiple that core cpu unit is 1,
Capacity is started to walk with 100GB, inside saves as 200MB startings.
If the user storage device meets the hardware requirement, user's storage that acquisition meets the hardware requirement is set
The user storage device is added corresponding storage according to the IP ownership place and set by standby IP ownership place, capacity and reliability
In standby cluster, according on the capacity and reliability based distribution corresponding PG to the user storage device.Wherein, the PG is
Placement groups are the logic storage units of Ceph, the mapping for carrying out tissue and position to the storage of object.Tool
For body, an object can only be mapped in a PG, and a PG can be then mapped in multiple storage devices, each storage
Equipment all carries a large amount of PG.
If the user storage device does not meet the hardware requirement, the addition for refusing the user storage device is asked
It asks.
In the specific embodiment of the invention, user storage device is added by the IP, capacity and reliability of user storage device
Enter cloud storage platform, solve the problems, such as the idle of user's local resource, improve the utilization rate of idle disk, while being cloud storage
Platform saves space cost.
Based on above-mentioned specific embodiment, Fig. 1 is a kind of structural schematic diagram of cloud storage platform of the specific embodiment of the invention,
As shown in Figure 1, a kind of cloud storage platform management method, further includes：S0, the IP ownership place for building the user storage device correspond to
The storage device cluster in region；The step S0 further comprises：S11, if the level where IP ownership place corresponding region
Further include other regions, and the load of the storage device cluster in any other regions meets preset loading condition,
Any user storage device is then exported from the storage device cluster in other any regions as remote backup equipment；It is no
Then, any server storage device is exported from main storage device cluster as remote backup equipment；The main storage device collection
Group is made of multiple server storage devices, and the server storage device is the storage device that service provider provides；S12, according to
The capacity and reliability of each user storage device are chosen several from the user storage device under IP ownership place corresponding region
A user storage device；S13 calculates the sum of the reliability of the remote backup equipment and several user storage devices, if institute
It states the sum of reliability and is not less than preset reliability cluster threshold value, then apply the remote backup equipment and several users
Storage device builds the storage device cluster of IP ownership place corresponding region.
Specifically, the storage of IP ownership place corresponding region is set based on server storage device and user storage device
Standby cluster is built, and method is as follows：
First, whether detect in the level where IP ownership place corresponding region further includes other regions：
If further including other regions in the level, the load of the storage device cluster in any other regions is judged
Whether preset loading condition is met：
If the load of the storage device cluster in any other regions meets the loading condition, from it is described it is any its
The storage device cluster output any user storage device in his region is set as the remote backup of IP ownership place corresponding region
It is standby；
If there is no other regions or IP ownership place corresponding region in the level where IP ownership place corresponding region
The storage device cluster in other regions in the level at place does not meet loading condition, then is exported from main storage device cluster
Remote backup equipment of any server storage device as IP ownership place corresponding region.The server storage device is
The storage device that the cloud storage platform provides, the main storage device cluster are made of multiple server storage devices.
Then, using the capacity of each user storage device and reliability as foundation, under IP ownership place corresponding region
Several user storage devices building for IP ownership place corresponding region storage device cluster is chosen in user storage device
It is vertical.
Finally, the remote backup equipment and several user storage devices obtained respectively in the first two steps is calculated
The sum of reliability, and the sum of judge the reliability and whether be less than preset reliability cluster threshold value：
If the sum of described reliability is not less than the reliability cluster threshold value, using above-mentioned remote backup equipment and several
A user storage device builds the storage device cluster of IP ownership place corresponding region.
In the specific embodiment of the invention, region storage device cluster is established by basis for estimation of reliability, is wrapped to realize
The management for having contained the cloud storage platform of user storage device is laid a good foundation.
Based on any of the above-described specific embodiment, Fig. 2 is a kind of cloud storage platform management method of the specific embodiment of the invention
Flow diagram, as shown in Fig. 2, a kind of cloud storage platform management method, further includes：S2 is set according to the storage in the region
At least one of reliability, network delay and the disk performance of each storage device, choose the storage in the region in standby cluster
The main storage device of device clusters.
Specifically, after user storage device being added in some region of storage device cluster, with depositing for the region
It is foundation to store up at least one of reliability, network delay and disk performance of each storage device in device clusters, to the area
The main storage device of the storage device cluster in domain is selected.
Wherein, each storage device in the storage device cluster in the region, including user storage device and server storage
Equipment.In the specific embodiment of the invention, the disk performance is using the IOPS values of the storage device as foundation.
Based on any of the above-described specific embodiment, a kind of cloud storage platform management method, if the memory device set in the region
In group, the reliability of any user storage device is less than preset minimum reliability threshold value, then deposits any user
Storage equipment is deleted from the storage device cluster.
Specifically, in the storage device cluster, ensure using minimum reliability threshold value each in the storage device cluster
Storage device can continue reliably to provide storage service to the user.If in storage device cluster, any user storage device
Reliability be less than the minimum reliability threshold value, then any user storage device is deleted from the storage device cluster
It removes.
Based on any of the above-described specific embodiment, a kind of cloud storage platform management method, the step S03 further includes：If institute
It states the sum of reliability and is less than preset reliability cluster threshold value, then export server is deposited from the main storage device cluster
Equipment is stored up, so that the sum of reliability being added after the server storage device is not less than preset reliability cluster threshold
Value builds the storage of current region using the server storage device, remote backup equipment and several described storage devices
Device clusters.
Specifically, the sum of the reliability of remote backup equipment and several user storage devices is calculated, and can described in judgement
Whether it is less than preset reliability cluster threshold value by the sum of degree：
If the sum of described reliability is not less than the reliability cluster threshold value, using above-mentioned remote backup equipment and several
The storage device cluster of a user storage device structure current region；
If the sum of described reliability is less than the reliability cluster threshold value, clothes are exported from the main storage device cluster
It is engaged in device storage device, and the sum of the reliability of the server storage device newly exported and described reliability is added.If update
The sum of reliability afterwards is not less than the reliability cluster threshold value, then the application remote backup equipment, several users store
The storage device cluster of equipment and the server storage device structure current region newly exported.
If the sum of updated reliability still is below the reliability cluster threshold value, continuation is set from the primary storage
Export server storage device in standby cluster, and by the sum of the reliability of the server storage device newly exported and the reliability
It is added, until the sum of updated reliability is not less than the reliability cluster threshold value.
In the specific embodiment of the invention, the side that cluster reliability is maintained by export server storage device is given
Method ensure that the reliability of storage device cluster, and guarantee is provided for the data safety of cloud storage platform.
Based on any of the above-described specific embodiment, a kind of cloud storage platform management method, in the storage device cluster at least
It is not less than the storage device of preset reliable apparatus threshold value including a reliability；The server storage device it is reliable
Degree immobilizes and is more than the reliable apparatus threshold value.
For example, the reliability cluster threshold value is 2.3, the reliability of the server storage device is 1；It is described reliably to set
Standby threshold value is 0.8, and the storage device that a reliability is not less than 0.8 is included at least in storage device cluster.
It is 0.8 to have 5 storage devices, two reliabilitys, and the reliability of the other three storage device is 0.5, then meeting
The reliability of condition is combined as 0.8+0.8+0.5+0.5=2.6>=2.3,0.8+0.5+0.5+0.5=2.3>=2.3.
Based on any of the above-described specific embodiment, a kind of cloud storage platform management method, if cloud storage cluster is added in application
User storage device meets preset hardware requirement, and the step S1 further includes：The user storage device is carried out suitable
Sequence reads and writes and random read-write, obtains the read or write speed and IOPS values of the user storage device.
Specifically, if the user storage device that cloud storage cluster is added in application meets preset hardware requirement, lead to
Cross to the user storage device carry out sequence read-write and random read-write come obtain the user storage device read or write speed and
IOPS values, and preserved the above-mentioned parameter measured as the disk assessment result of the user storage device.
Based on any of the above-described specific embodiment, a kind of cloud storage platform management method further includes：S3, waiting are preset
Time after, update the reliability of each storage device in the storage device cluster in the region, execute step S2.
Specifically, for some region of user storage device, when in use between reache a certain level after, the user deposits
The reliability of storage equipment also can be promoted gradually.Over time, meet in the storage device cluster in region and described reliably set
The storage device quantity of standby threshold value can also increase, and the newly-increased storage device for meeting the reliable apparatus threshold value may be than current
Main storage device performance is more preferably.
Thus, after the main storage device in the storage device cluster to region is chosen, the preset time is waited for, more
The reliability of each storage device in the storage device cluster in the new region, and based on the reliable of updated each storage device
Degree, again chooses the main storage device in the storage device cluster.
In the specific embodiment of the invention, the reliability by updating storage equipment, which updates, chooses main storage device so that deposits
Main storage device performance in storage device clusters maintains optimum state, and sound assurance is provided for the storage of critical data.
Based on any of the above-described specific embodiment, a kind of cloud storage platform management method is set when according to the storage in the region
Reliability, network delay and the disk performance of each storage device, choose the master of the storage device cluster in the region in standby cluster
When storage device, the step S2 further comprises：S21 extracts reliability in the storage device cluster in the region and is not less than
Each storage device of preset reliable apparatus threshold value；S22 is respectively stored described in the form calculus using random sampling point ping packets
The network delay of equipment is converted each storage of corresponding acquisition by percentage and is set on the basis of lowest latency in the network delay
Standby network delay score；It is converted using highest IOPS values in the IOPS values of each storage device as benchmark by percentage
The corresponding disk performance score for obtaining each storage device；S23, by the network delay score and disk performance of each storage device
Comprehensive score of the sum of the score as each storage device, extracting the highest storage device of comprehensive score in each storage device is
The main storage device of the storage device cluster in the region.
Specifically, after user storage device being added in some region of storage device cluster, with depositing for the region
Reliability, network delay and the disk performance for storing up each storage device in device clusters are foundation, to the storage device in the region
The main storage device of cluster is selected, and is further comprised：
First, according to preset reliable apparatus threshold value, reliability is chosen in the storage device cluster in the region not
Less than each storage device of the reliable apparatus threshold value, storage device described herein includes server storage device and user's storage
Equipment.
Then, the network delay score and disk performance score of each storage device chosen in above-mentioned steps are calculated separately.
Wherein, the network delay of each storage device is obtained by the form calculus of random sampling point ping packets, then, with above-mentioned
Most young value in each storage device in network delay, i.e. lowest latency set each storage as benchmark, by percentage
Standby network delay score is converted；
Meanwhile the IOPS values of each storage device are obtained, the IOPS values of wherein user storage device are stored from the user
It is extracted in the disk assessment result of equipment.With the optimal value of IOPS in above-mentioned each storage device, i.e., on the basis of maximum IOPS values,
The disk performance score of each storage device is converted by percentage
Finally, by the sum of the network delay score of each storage device and disk performance score as each storage device
Comprehensive score extracts the storage device cluster that the highest storage device of comprehensive score in each storage device is the region
Main storage device.
The specific embodiment of the invention gives the side that main storage device is chosen according to reliability, network delay and disk performance
Method provides guarantee for the storage of critical data.
Based on any of the above-described specific embodiment, a kind of cloud storage platform management method, the user storage device it is reliable
Degree is calculated by following formula and is obtained：
Reliability=a reference value × disk life value × availability × length of delay+disk performance
Wherein, the disk life value ∈ [0,1], the availability ∈ [0,1], the length of delay ∈ [0,1].
For example, the reliability ∈ [0,1], the reliability of server storage device remain 1；
If user storage device is that the cloud storage cluster is added for the first time, a reference value is 0.5；Otherwise, the base
Quasi- value is 0.2；
If the disk usage time of the user storage device is less than 2 years, the disk life value is 1；Otherwise, institute
It states disk usage time beyond 2 years parts often to increase by 3 months, the disk life value is multiplied by 0.95；The disk life value
The disk life value is introduced the calculating of reliability, can reduce what the overage user storage device used was brought by ∈ [0,1]
Destabilizing factor；
The availability be nearest one week of the user storage device can service minutes and nearest one week total the number of minutes
Quotient；The calculating that the availability is introduced to reliability, for characterizing holding for the user storage device in reliability
Continuous service rate.If the time of the cloud storage platform is added less than one week, by the availability generation in the user storage device
Enter ratio conversion factor, the ratio conversion factor is as follows：
Wherein, x is the availability before correction.
If the delay of the user storage device is less than 100ms, the length of delay is 1；Otherwise, the delay exceeds
The part of 100ms often increases 10ms, and the length of delay is multiplied by 0.95；The length of delay ∈ [0,1].
The disk performance is the IOPS values and the difference of IOPS benchmark and the product of proportionality coefficient of the user storage device.
For example, IOPS benchmark are 500, proportionality coefficient 0.004, IOPS values are 1000, then disk performance is 0.2.The disk performance
∈[-0.2,0.2]。
In the specific embodiment of the invention, reliability is related to disk life value, availability, length of delay and disk performance, has
Help the service quality of more accurate comprehensive reaction storage device.
Carry out following with a kind of cloud storage platform management method proposed by the present invention, the present invention is applied in order to better understand
Example, and the present invention is not only limited to following example.
Example one：
To realize the management of cloud storage platform, it is proposed that a kind of reliability mechanism.The reliability mechanism is used for assessing
The credible event for the storage device that family provides ensures that the data of cloud storage platform stablize available state in one.
The measurement facility of the reliability mechanism is reliability and marker bit osd_type, and the value range of reliability is
[0.2,1].There are three types of osd_type, and 0 indicates server storage device, is the hard disk provided by service provider；1 indicates reliably to set
It is standby, belong to hard disk that user provides and relatively reliable；2 indicate conventional equipment, belong to user's offer but reliability is lower
Equipment.The reliability for the hard disk that cloud storage platform service quotient provides all is 1, osd_type 0, indicates that these storage devices are opposite
In user provide storage device be metastable equipment.The critical data of cloud platform can be all stored in.When reliability is less than
When 0.2, user storage device can be removed out cluster.The wherein sequence number of cluster meeting recording disk, and record storage is being taken
It is engaged in device storage device cluster, it is 0.2 that the storage device being removed, which rejoins rear reliability,.Reliability is more than or equal to 0.8
Storage device is known as reliable memory equipment (osd_type marker bits are 1).In the corresponding reliability set of one data, it is necessary to extremely
It is few to belong to reliable memory equipment there are one storage device.In storage device, the value of osd_type is exactly reliably to set less than or equal to 1
It is standby.
Example two：
For user storage device, encryption and decryption is carried out when reading and writing data, data is prevented to be stolen.If user storage device
Reliability be less than 0.2, i.e., rejected from the cluster, then the data in the user storage device will be by automatic clear
Sky, and forbid being added in 1 day.N-th is suggested cluster, needs equal 2N-1Number of days can just rejoin the cloud storage platform.
In addition, limiting user terminal, prevent it from accessing described in corresponding user storage device and the user storage device
Data in storage device cluster.
Finally, the present processes are only preferable embodiment, are not intended to limit the scope of the present invention.It is all
Within the spirit and principles in the present invention, any modification, equivalent replacement, improvement and so on should be included in the protection of the present invention
Within the scope of.
Claims (10)
1. a kind of cloud storage platform management method, which is characterized in that including：
S1, if the user storage device that cloud storage cluster is added in application meets preset hardware requirement, by the user
Storage device is added in the storage device cluster of IP ownership place corresponding region of the user storage device, and according to the user
On capacity and the reliability based distribution corresponding PG to the user storage device of storage device.
2. according to the method described in claim 1, it is characterized in that, further including：
S0 builds the storage device cluster of the IP ownership place corresponding region of the user storage device；
The step S0 further comprises：
S01, if the level where IP ownership place corresponding region further includes other regions, and any other regions are deposited
The load of storage device clusters meets preset loading condition, then defeated from the storage device cluster in other any regions
Go out any user storage device as remote backup equipment；Otherwise, any server storage is exported from main storage device cluster
Equipment is as remote backup equipment；The main storage device cluster is made of multiple server storage devices, and the server is deposited
Storage equipment is the storage device that service provider provides；
S02 is set according to the capacity and reliability of each user storage device from user's storage under IP ownership place corresponding region
Standby middle several user storage devices of selection；
S03 calculates the sum of the reliability of the remote backup equipment and several user storage devices, if the sum of described reliability
Not less than preset reliability cluster threshold value, then the application remote backup equipment and several user storage devices are built
The storage device cluster of IP ownership place corresponding region.
3. according to the method described in claim 1, it is characterized in that, further including：
S2, according in the reliability of each storage device, network delay and disk performance in the storage device cluster in the region
At least one chooses the main storage device of the storage device cluster in the region.
4. according to the method described in claim 1, it is characterized in that, further including：
If in the storage device cluster in the region, the reliability of any user storage device is less than preset lowest reliable
Threshold value is spent, then is deleted any user storage device from the storage device cluster.
5. according to the method described in claim 2, it is characterized in that, the step S03 further includes：
If the sum of described reliability is less than preset reliability cluster threshold value, exported from the main storage device cluster
Server storage device, so that the sum of reliability being added after the server storage device is reliable not less than preset
Cluster threshold value is spent, works as proparea using the server storage device, remote backup equipment and several described storage devices structure
The storage device cluster in domain.
6. according to the method described in claim 2, it is characterized in that, including at least a reliability in the storage device cluster
Not less than the storage device of preset reliable apparatus threshold value；The reliability of the server storage device immobilizes and greatly
In the reliable apparatus threshold value.
7. according to the method described in claim 2, it is characterized in that, if the user storage device symbol of cloud storage cluster is added in application
Preset hardware requirement is closed, the step S1 further includes：
To the user storage device carry out sequence read-write and random read-write, obtain the user storage device read or write speed and
IOPS values.
8. according to the method described in claim 3, it is characterized in that, further including：
S3 after waiting for preset duration, updates the reliability of each storage device in the storage device cluster in the region, executes at least
Step S2.
9. the method according to claim 3 or 8, which is characterized in that when each in the storage device cluster according to the region
Reliability, network delay and the disk performance of storage device, when choosing the main storage device of storage device cluster in the region,
The step S2 further comprises：
S21 extracts reliability respectively depositing not less than preset reliable apparatus threshold value in the storage device cluster in the region
Store up equipment；
S22, the network delay of each storage device described in the form calculus using random sampling point ping packets, in the network delay
On the basis of lowest latency, the corresponding network delay score for obtaining each storage device is converted by percentage；
Using highest IOPS values in the IOPS values of each storage device as benchmark, corresponding each storage of acquisition is converted by percentage
The disk performance score of equipment；
S23, the synthesis by the sum of the network delay score of each storage device and disk performance score as each storage device
Score extracts the main memory for the storage device cluster that the highest storage device of comprehensive score in each storage device is the region
Store up equipment.
10. the method according to any claim in claim 1 to 8, which is characterized in that the user storage device
Reliability is calculated by following formula and is obtained：
Reliability=a reference value × disk life value × availability × length of delay+disk performance
Wherein, the disk life value ∈ [0,1], the availability ∈ [0,1], the length of delay ∈ [0,1].
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810041455.6A CN108366094B (en) | 2018-01-16 | 2018-01-16 | Cloud storage platform management method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810041455.6A CN108366094B (en) | 2018-01-16 | 2018-01-16 | Cloud storage platform management method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108366094A true CN108366094A (en) | 2018-08-03 |
| CN108366094B CN108366094B (en) | 2021-05-07 |
Family
ID=63006397
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810041455.6A Active CN108366094B (en) | 2018-01-16 | 2018-01-16 | Cloud storage platform management method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108366094B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109669642A (en) * | 2018-12-24 | 2019-04-23 | 广东浪潮大数据研究有限公司 | Node Adding Way, system, device and the readable storage medium storing program for executing of storage system |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101902498A (en) * | 2010-07-02 | 2010-12-01 | 广州鼎甲计算机科技有限公司 | Network technology based storage cloud backup method |
| CN102082830A (en) * | 2011-01-18 | 2011-06-01 | 浙江大学 | Unstable network-oriented distributed file storage method based on quality perception |
| CN102523259A (en) * | 2011-12-02 | 2012-06-27 | 浪潮集团有限公司 | Method for accessing personal computer (PC) to universal serial bus (USB) interface cloud terminal |
| CN103200168A (en) * | 2013-01-21 | 2013-07-10 | 马国强 | Method for achieving interaction with network hard disk device through wide area network and network hard disk device |
| CN103246546A (en) * | 2013-05-07 | 2013-08-14 | 山东大学 | Open power system numerical simulation system and method based on cloud technology |
| CN104380261A (en) * | 2012-05-16 | 2015-02-25 | 苹果公司 | Locally backed cloud-based storage |
| CN104883393A (en) * | 2015-04-28 | 2015-09-02 | 深圳市美贝壳科技有限公司 | Household cloud storage method |
| US20160127339A1 (en) * | 2014-10-30 | 2016-05-05 | Lenovo (Singapore) Pte. Ltd. | Aggregate service with enhanced cloud device management |
| CN106230963A (en) * | 2016-08-11 | 2016-12-14 | 江苏方天电力技术有限公司 | The software definition storage method of the privately owned cloud platform of electrically-based system |
| CN106990923A (en) * | 2017-03-30 | 2017-07-28 | 武汉大学 | A kind of Dropbox constructing apparatus and method based on personal memory device |
| CN107171904A (en) * | 2017-05-03 | 2017-09-15 | 易展成 | A kind of cloud service provides method and device |
-
2018
- 2018-01-16 CN CN201810041455.6A patent/CN108366094B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101902498A (en) * | 2010-07-02 | 2010-12-01 | 广州鼎甲计算机科技有限公司 | Network technology based storage cloud backup method |
| CN102082830A (en) * | 2011-01-18 | 2011-06-01 | 浙江大学 | Unstable network-oriented distributed file storage method based on quality perception |
| CN102523259A (en) * | 2011-12-02 | 2012-06-27 | 浪潮集团有限公司 | Method for accessing personal computer (PC) to universal serial bus (USB) interface cloud terminal |
| CN104380261A (en) * | 2012-05-16 | 2015-02-25 | 苹果公司 | Locally backed cloud-based storage |
| CN103200168A (en) * | 2013-01-21 | 2013-07-10 | 马国强 | Method for achieving interaction with network hard disk device through wide area network and network hard disk device |
| CN103246546A (en) * | 2013-05-07 | 2013-08-14 | 山东大学 | Open power system numerical simulation system and method based on cloud technology |
| US20160127339A1 (en) * | 2014-10-30 | 2016-05-05 | Lenovo (Singapore) Pte. Ltd. | Aggregate service with enhanced cloud device management |
| CN104883393A (en) * | 2015-04-28 | 2015-09-02 | 深圳市美贝壳科技有限公司 | Household cloud storage method |
| CN106230963A (en) * | 2016-08-11 | 2016-12-14 | 江苏方天电力技术有限公司 | The software definition storage method of the privately owned cloud platform of electrically-based system |
| CN106990923A (en) * | 2017-03-30 | 2017-07-28 | 武汉大学 | A kind of Dropbox constructing apparatus and method based on personal memory device |
| CN107171904A (en) * | 2017-05-03 | 2017-09-15 | 易展成 | A kind of cloud service provides method and device |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109669642A (en) * | 2018-12-24 | 2019-04-23 | 广东浪潮大数据研究有限公司 | Node Adding Way, system, device and the readable storage medium storing program for executing of storage system |
| CN109669642B (en) * | 2018-12-24 | 2022-06-10 | 广东浪潮大数据研究有限公司 | Node joining method, system and device of storage system and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108366094B (en) | 2021-05-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8805902B2 (en) | Managing snapshot storage pools | |
| US20190294338A1 (en) | Selecting pages implementing leaf nodes and internal nodes of a data set index for reuse | |
| CN107168884A (en) | The control method of storage system, information processing system and nonvolatile memory | |
| CN107168885A (en) | The control method of storage system, information processing system and nonvolatile memory | |
| CN105100146B (en) | Date storage method, apparatus and system | |
| CN107168639A (en) | The control method of storage system, information processing system and nonvolatile memory | |
| CN107168640A (en) | The control method of storage system, information processing system and nonvolatile memory | |
| CN103412822B (en) | Operation Nonvolatile memory and the method and relevant apparatus of data manipulation | |
| KR20120102664A (en) | Allocating storage memory based on future use estimates | |
| US20100023566A1 (en) | Method for Allocating Logical Unit Inside Storage System to Logical Volume and Storage Controller | |
| CN110222013B (en) | Method, system, equipment and storage medium for determining cluster storage capacity | |
| CN107924291A (en) | Storage system | |
| CN109428913A (en) | A kind of storage expansion method and device | |
| CN109117088A (en) | A kind of data processing method and system | |
| CN107623732A (en) | A kind of date storage method based on cloud platform, device, equipment and storage medium | |
| CN109471843A (en) | A kind of metadata cache method, system and relevant apparatus | |
| CN115878595A (en) | FTL-based intelligent Internet of things table data migration method and storage medium | |
| CN108710686A (en) | A kind of date storage method, device, storage medium and terminal | |
| CN113760187B (en) | Method, system, terminal and storage medium for generating deduplication IO thread | |
| CN108366094A (en) | A kind of cloud storage platform management method | |
| CN109299043A (en) | The big file delet method of distributed cluster system, device, equipment and storage medium | |
| CN103761194B (en) | A kind of EMS memory management process and device | |
| CN109960474A (en) | Data-updating method, device, equipment and medium based on automatic simplify configuration | |
| CN105138278B (en) | A kind of the naked of intelligent substation network message takes inventory method for storing | |
| CN110531934A (en) | Date storage method and device based on distributed system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |