Quality of service differentiated cloud storage

• US 8,612,284 B1
• Filed: 11/09/2011
• Issued: 12/17/2013
• Est. Priority Date: 11/09/2011
• Status: Active Grant

• Current Assignee: Virtuozzo International GmbH
• Original Assignee: Parallels Holdings Limited (HNA Technology Co. Ltd.)
• Inventors: Sharif, Amir, Korotaev, Kirill
• Primary Examiner(s): Flynn, Kevin
• Assistant Examiner(s): TALLMAN, BRIAN A
• Application Number:
  US13/292,986
• Time in Patent Office: 769 Days
• Field of Search: 705/7.35, 705/400, 705/80, 707/812
• US Class Current: 705/7.35
• CPC Class Codes: G06F 3/0611 in relation to response time
• CPC Class Codes: G06F 3/064 Management of blocks
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
• CPC Class Codes: H04L 67/1097 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]

Abstract

A system, method and computer program product for managing the quality of service differentiated cloud storage. The cloud storage (and its services) is highly distributed. An overall speed of the storage is calculated based on storage speed and a traffic speed of the connection to the particular cloud storage. Data access expenses are calculated. Then, for each of the portions of the cloud storage the associated speed and traffic costs are calculated, so the user can select a level of data access guarantee (DAG) for his particular needs. In order to maintain the desired overall speed of the data access, geo-targeting with the cloud storage system is implemented, where the same data is retrieved from different locations based on optimal data movement costs.

Claims

1. A system for managing cloud storage based on quality of service, comprising:

a data file divided into data chunks;

a plurality of cloud storage devices configured to store geographically distributed replicas of the data chunks in a continuous manner, wherein the cloud storage devices are further configured to be accessed by client machines via plurality of connections having traffic differentiated by speed and price, and wherein the cloud storage devices are further configured to be differentiated by price and speed of data retrieval of hardware storage elements of the cloud storage devices,

and wherein the cloud storage devices are further configured to be controlled by storage plan parameters, wherein the storage plan parameters include an overall speed of data retrieval,

wherein the overall speed of data retrieval is a product of storage data retrieval speed and traffic speed of a corresponding connection between a client machine and one of the cloud storage devices, and wherein a data access guarantee (DAG) is based on the overall speed of data retrieval, requiring at least some client'"'"'s data to move from one cloud storage device to a closer cloud storage device as the client machine moves geographically,

wherein the storage plan parameters for speed, reliability and price are based on information provided to the client machine by the system,

wherein the DAG is calculated as K1*(cloud storage device speed)*(traffic connection speed)+K2*(1−

(cloud storage device reliability)*(traffic connection reliability)), K1 and K2 being constants, and

wherein the storage plan parameters are dynamically configurable at runtime by the client machine.

2. The system of claim 1, wherein the storage plan parameters include a price of data access, wherein the price of data access is based on a cloud storage price multiplied by a corresponding connection traffic price.

3. The system of claim 1, wherein the storage plan parameters include a price of data access, wherein the price of data access is based on reliability values of each cloud storage device and a corresponding client machine connection.

4. The system of claim 1, further comprising a plurality of metadata servers (MDSs) configured to provide metadata required to access data within the cloud storage devices.

5. The system of claim 4, wherein the metadata provided by the MDSs includes any of:

file name;

data indicating how the files are split into the data chunks;

data chunk residence cloud server;

data chunk size; and

data chunk location.

6. A method for managing cloud storage based on quality of service, the method comprising:

creating a plurality of interconnected cloud storage devices;

creating at least one redundant storage to continuously store data replicas;

determining a speed of hardware storage elements for each of the cloud storage devices;

measuring connection traffic speed;

on a server device, calculating prices of the cloud storage devices based on the speed of data retrieval from the hardware storage elements and connection traffic speed;

estimating reliability of the cloud storage devices;

calculating anytime data access guarantee (DAG) based on the cloud storage device reliability and traffic connection reliability;

calculating costs of moving data between the cloud storage devices and the redundant storage to achieve a calculated DAG; and

transmitting to a client machine, over a network, a least expensive cost of moving the data to a geographic location closer to the client machine,

wherein the DAG is calculated as K1*(cloud storage device speed)*(traffic connection speed)+K2*(1−

(cloud storage device reliability)*(traffic connection reliability)), K1 and K2 being constants.

7. The method of claim 6, wherein a client selects desired cloud service based on the DAG.

8. The method of claim 6, further comprising:

holding the cloud storage device speed multiplied by the connection traffic speed to a constant value.

9. The method of claim 6 further comprising:

determining whether a cloud service provider has a positive margin when the costs of moving data between the cloud storage devices and the redundant storage are larger than zero.

10. The method of claim 6, wherein the cloud storage devices are located in distributed geographic locations.

11. The method of claim 10 further comprising:

optimizing the costs of moving data between the cloud storage devices and the redundant storage providing efficient geo-targeting of data.

12. The method of claim 6 further comprising:

optimizing the costs of moving data between the cloud storage devices and the redundant storage providing optimization of network.

13. The method of claim 6 further comprising:

optimizing the costs of moving data between the cloud storage devices and the redundant storage providing optimization of bulk transfer rate.

14. A system for managing cloud storage based on quality of service, the system comprising:

a processor;

a memory coupled to the processor;

a computer program logic stored in the memory and executed on the processor, the computer program logic implementing the steps of;

creating a plurality of interconnected cloud storage devices;

creating at least one redundant storage for continuously storing data replicas;

determining a speed of hardware storage elements for each of the cloud storage devices;

measuring connection traffic speed;

calculating prices of the cloud storage devices based on the speed of data retrieval from the hardware storage elements and connection traffic speed;

estimating reliability of the cloud storage devices;

calculating anytime data access guarantee (DAG) based on the cloud storage device reliability and traffic connection reliability;

calculating costs of moving data between the cloud storage devices and the redundant storage for a calculated DAG; and

transmitting to a client machine, over a network, a least expensive cost of moving the data to a geographic location closer to the client machine,

wherein the DAG is calculated as K1*(cloud storage device speed)*(traffic connection speed)+K2*(1−

(cloud storage device reliability)*(traffic connection reliability)), K1 and K2 being constants.

15. A system for managing cloud storage based on data access guarantees (DAG), comprising:

a plurality of cloud storage devices in a cloud configured to store geographically distributed data, using redundancy and continuous replication, wherein the cloud storage devices are further configured to be accessed by client computers via plurality of connections, and wherein the client computers are geographically distributed and move geographically over time,

wherein the cloud storage devices are further configured to have speed of data retrieval parameters of hardware storage elements of the cloud storage devices and storage reliability parameters, and the plurality of connections have data transmission speed and access reliability parameters,

wherein the cloud storage devices are further configured to be controlled by storage plan parameters, wherein the storage plan parameters are part of a proposed storage plan of a plurality of storage plans, wherein the proposed storage plan includes an estimated cost value based on allocation of the cloud storage devices and speed of transmitting channels from the cloud storage devices to a client computer,

wherein the proposed storage plan includes DAG, wherein the DAG include criteria of at least the speed of data retrieval parameters, the data transmission speed parameters and the access reliability parameters for each session when the client computer accesses its data on the cloud,

wherein the DAG is based on a level of the geographical distribution of cloud storage device locations of replicated data,

wherein the DAG corresponding to the proposed storage plan is dynamically alterable, and the system provides dynamic redistribution of client data inside a cloud to different storage devices as the client computer'"'"'s geographic location changes, to satisfy the DAG,

and wherein the DAG is calculated as K1*(cloud storage device speed)*(traffic connection speed)+K2*(1−

(cloud storage device reliability)*(traffic connection reliability)), K1 and K2 being constants.

16. The system of claim 15, wherein at least one of the following are configured to geographically redundantly distribute the data to achieve the level of the geographical distribution:

different servers;

different server racks; and

different stand of racks of servers.

17. The system of claim 15, wherein at least one of the following are configured to geographically redundantly distribute the data to achieve the level of the geographical distribution:

different halls in a data center; and

different server centers in a single building.

18. The system of claim 15, wherein at least one of the following are configured to geographically redundantly distribute the data to achieve the level of the geographical distribution:

different server centers in the same country; and

different server centers in different countries.

19. The system of claim 15, wherein the data includes financial data.

20. The system of claim 19, wherein the DAG is based on the financial data including highest available reliability parameters with the financial data.

21. The system of claim 15, wherein the data includes entertainment data, wherein the DAG is based on the entertainment data including highest available data transmission speed parameters with the entertainment data.