Distributed object storage service quality assurance method and system based on lease

By adopting a lease-based flow control method in a distributed storage system, the scalability and reliability problems of distributed storage systems in multi-data center deployment are solved, and high-performance and high-scalability distribution is achieved storage system.

CN112383486BActive Publication Date: 2024-12-03TOYOU FEIJI ELECTRONICS
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202011223479.7
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-11-05
Publication Date
2024-12-03
Estimated Expiration
2040-11-05

AI Technical Summary

Technical Problem

Existing distributed storage systems have scalability and reliability problems in traffic control, especially in multi-data center deployment environments, where IO request processing requires access to third-party entities, resulting in the impact of system scalability and reliability.

Method used

Traffic control is performed using a lease-based method, and the traffic rate is obtained by timed update of the lease, and independently control it in the distributed storage gateway, avoiding access to the lease management center, and using asynchronous timing tasks to process status information reporting and lease updates.

Benefits of technology

The performance and scalability of distributed storage systems are achieved without the impact of traffic control, reducing the pressure on lease management centers, and maintaining high availability of systems in large-scale multi-data center deployments.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112383486B_ABST
    Figure CN112383486B_ABST
Patent Text Reader

Abstract

The present invention provides a distributed object storage service quality assurance method and system based on lease. The method includes: each object storage gateway independently performs flow control according to the parameters in the lease, obtains the flow rate allocated to the node within a period of time by regularly updating the lease, and each service node uses a part of the total flow rate within the lease time range to perform flow control, thereby achieving the effect of overall flow control. The distributed object storage service quality assurance method and system based on lease provided by the present invention can reduce the impact of flow control on the scalability and reliability of distributed object storage.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of distributed service quality assurance, and in particular to a distributed object storage service quality assurance method and system based on a lease. Background Art

[0002] With the rapid development of cloud computing and big data, traditional centralized storage is increasingly unable to meet the needs of cloud computing and big data applications. Distributed storage systems have become an important part of cloud computing platforms due to their flexible expansion capabilities. Cloud computing environments usually use a multi-tenant approach to allocate and manage resources. Therefore, how to allocate and manage limited bandwidth and request processing capabilities to ensure the service quality of different tenants is an important issue that distributed storage systems need to solve.

[0003] For single-node service environments, the industry currently generally uses relatively mature token bucket and leaky bucket algorithms for flow control. There are usually two ways to control flow in multi-node distributed storage systems:

[0004] Method 1: Introduce a centralized proxy gateway to control requests and traffic at the proxy gateway. The disadvantage of this solution is that the centralized proxy gateway limits the scalability of the distributed object storage system and introduces a single point problem.

[0005] Method 2: Introduce a cache system such as Redis to store the status information of flow control in the cache, and multiple nodes can implement flow control by sharing the cache status. In this solution, each IO request to access the storage will update the status information in the cache. The performance and reliability of the cache system restrict the scalability of integrated distributed storage to a certain extent.

[0006] None of the above solutions decouples the IO request and flow control of distributed storage, so that the IO request processing process still needs to access third-party entities, affecting the scalability and reliability of the system. In particular, for large-scale multi-data center deployed distributed object storage systems, the above methods cannot support them well because of the large network delay between data centers. Summary of the invention

[0007] The technical problem to be solved by the present invention is to provide a distributed object storage service quality assurance method and system based on a lease, which can reduce the impact of flow control on the scalability and reliability of distributed object storage.

[0008] To solve the above technical problems, the present invention provides a distributed object storage service quality assurance method based on a lease, which is applied to an object storage gateway, including: independently performing flow control according to parameters in the lease, obtaining the flow rate allocated to the node within a period of time by regularly updating the lease, and each service node using a part of the total flow rate within the lease time range for flow control, thereby achieving the effect of overall flow control.

[0009] In some implementations, the method further includes: regularly reporting IO statistics information to a lease management center.

[0010] In some implementations, the further step includes: the lease management center calculating a traffic rate control value allocated to each object storage gateway according to the IO status.

[0011] In some implementations, the flow control lease information is obtained from a lease management center.

[0012] In some implementations, two leases are maintained and switched according to the validity period of the leases to maintain the continuity of available leases.

[0013] In some implementations, flow control is performed according to parameters in a flow control lease.

[0014] In addition, the present invention also provides a lease-based distributed object storage service quality assurance system, the system comprising: one or more processors; a storage device for storing one or more programs, when the one or more programs are executed by the one or more processors, the one or more processors implement the lease-based distributed object storage service quality assurance method described above.

[0015] After adopting such a design, the present invention has at least the following advantages:

[0016] Status information reporting and lease renewal are handled by asynchronous scheduled tasks, which are independent of the storage IO processing process. That is, the storage IO processing process does not need to access the lease management center, and the throughput of the distributed storage system is not limited by third-party systems. The performance and scalability of the system are not affected by flow control.

[0017] Since the update frequency of leases and status does not depend on the IO rate of the system, the pressure on the lease management center will not increase due to the increase in the IO pressure of the storage system. The lease management center can be managed by relatively mature configuration management centers such as zookeeper and ectd, or it can directly save global lease information to distributed storage for centralized management.

[0018] Because the distributed storage gateway and the lease management center in the present invention use a timed communication method and have a certain tolerance for network delays, it is still applicable to distributed object storage systems deployed in large-scale multiple data centers. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The above is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0020] Figure 1 is a system architecture diagram provided by an embodiment of the present invention;

[0021] Figure 2 This is a flow chart of IO status reporting and lease acquisition provided by an embodiment of the present invention;

[0022] Figure 3 It is a flowchart of object storage gateway lease switching provided by an embodiment of the present invention;

[0023] Figure 4 It is a system architecture diagram of a lease-based distributed object storage service quality assurance system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0024] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0025] The relevant components and system architecture of the present invention are as follows Figure 1 As shown, the client uses the object storage service by accessing the interface of the distributed storage gateway. The distributed storage gateway is usually deployed with multiple nodes to improve system performance and reliability. The client selects different storage gateways to access according to the strategy. The lease management center is an independent entity introduced by the present invention, which consists of three modules: the QoS configuration management module is used to manage QoS related configuration parameters, such as: the maximum number of requests per second, the number of bytes read and written per second, etc.; the IO statistics management module is responsible for the management of the IO statistical information of each distributed storage gateway, and the IO load status of each storage gateway node is evaluated based on this information; the QoS lease management module is mainly responsible for the allocation and management of the QoS lease of each storage gateway.

[0026] The process of IO status reporting and lease acquisition is as follows Figure 2As shown, the distributed storage gateway counts the IOs it processes according to certain rules, and regularly reports these statistics and the expiration time of the latest lease it saves to the lease management center. The lease management center compares the lease expiration time in the request with the lease saved in the QoS lease management module. If the lease expiration time in the request is less than the lease expiration time in the lease management center, the lease is returned to the storage gateway in the return message. The storage gateway replaces the locally saved lease with an earlier expiration time with the latest acquired lease.

[0027] The distributed storage gateway lease switching process is as follows: Figure 3 The distributed storage gateway locally stores two leases, one is the lease that is currently in effect, and the other is the lease that is about to be used or has expired. Figure 2 The lease timing acquisition process shown can keep the distributed storage gateway continuously acquiring the latest lease. During the IO request process, the flow control parameters in the current lease are obtained. If the current lease is still valid, the current lease is used directly. If the current lease is expired, check whether another lease is valid. If valid, update the current lease's identifier to this lease and return it. Otherwise, generate a temporary lease and use it temporarily. After the lease is updated, switch to the allocated lease.

[0028] The lease management center calculates the traffic rate control value that can be used by each gateway node in the next period of time before the previous lease expires based on the IO statistics reported by each distributed storage gateway. The specific calculation method and parameters are as follows:

[0029] Assume that the waiting time of IO request is L, the length of statistical list is n, and the weighted sum of IO statistical information reported by each gateway node is obtained The bandwidth and processing capacity weight value W of the storage gateway, the number of storage gateways m, the guaranteed traffic proportion u, and the total traffic rate control value Limit are calculated by the following formula: the traffic rate control value γ allocated to each storage gateway node i :

[0030]

[0031] According to the above formula, the lease management center regularly calculates the lease of each storage gateway node for the next period of time. When the node reports the IO status, the latest lease information is returned to the storage gateway node. For the specific flow, see Figure 2 .

[0032] Figure 4 The structure of the distributed object storage service quality assurance system based on lease is shown. Figure 4For example, the lease-based distributed object storage service quality assurance system 400 can be used to act as an object storage gateway in a distributed service quality assurance system. As described herein, the lease-based distributed object storage service quality assurance system 400 can be used to implement lease-based traffic control functions in a distributed service quality assurance system. The lease-based distributed object storage service quality assurance system 400 can be implemented in a single node, or the functions of the lease-based distributed object storage service quality assurance system 400 can be implemented in multiple nodes in a network. Those skilled in the art should appreciate that the term lease-based distributed object storage service quality assurance system includes devices in a broad sense, Figure 4 The lease-based distributed object storage service quality assurance system 400 shown in the specification is only one example. The lease-based distributed object storage service quality assurance system 400 is included for clarity of description and is not intended to limit the application of the present invention to a specific lease-based distributed object storage service quality assurance system embodiment or a certain class of lease-based distributed object storage service quality assurance system embodiments. At least some of the features / methods described in the present invention can be implemented in a network device or component, such as a lease-based distributed object storage service quality assurance system 400. For example, the features / methods in the present invention can be implemented in hardware, firmware and / or software installed and run on hardware. The lease-based distributed object storage service quality assurance system 400 can be any device that processes, stores and / or forwards data frames through a network, such as a server, a client, a data source, etc. Figure 4 As shown, the lease-based distributed object storage service quality assurance system 400 may include a transceiver (Tx / Rx) 410, which may be a transmitter, a receiver, or a combination thereof. Tx / Rx 410 may be coupled to multiple ports 450 (e.g., an upstream interface and / or a downstream interface) for sending and / or receiving frames from other nodes. A processor 430 may be coupled to Tx / Rx 410 to process frames and / or determine which nodes to send frames to. The processor 430 may include one or more multi-core processors and / or a memory device 432, which may be used as a data storage device, a buffer, etc. The processor 430 may be implemented as a general-purpose processor, or may be part of one or more application-specific integrated circuits (ASICs) and / or digital signal processors (DSPs).

[0033] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Those skilled in the art may make some simple modifications, equivalent changes or modifications using the technical contents disclosed above, which all fall within the protection scope of the present invention.

Claims

1. A distributed object storage service quality assurance method based on lease, applied to an object storage gateway, characterized in that: include: Independently perform flow control according to the parameters in the lease. The lease is updated regularly to obtain the flow rate allocated to the node within a period of time. Each service node uses a part of the total flow rate within the lease time range for flow control, thereby achieving the effect of overall flow control. Regularly report IO statistics to the lease management center; The lease management center calculates the traffic rate control value allocated to each object storage gateway based on the IO status, including: The lease management center calculates the traffic rate control value that can be used by each gateway node in the next period of time before the previous lease expires based on the IO statistics reported by each distributed storage gateway. The specific calculation method and parameters are as follows: Assume that the waiting time of IO request is L, the length of statistical list is n, and the weighted sum of IO statistical information reported by each gateway node is obtained The bandwidth and processing capacity weight value W of the storage gateway, the number of storage gateways m, the guaranteed traffic proportion u, and the total traffic rate control value Limit are calculated by the following formula: the traffic rate control value γ allocated to each storage gateway node i : The lease management center periodically calculates the lease of each storage gateway node for the next period of time, and returns the latest lease information to the storage gateway node when the node reports the IO status.

2. The distributed object storage service quality assurance method based on lease according to claim 1 is characterized in that: Obtain flow control lease information from the lease management center.

3. The distributed object storage service quality assurance method based on lease according to claim 1 is characterized in that: Maintain two leases and switch them according to the validity period of the leases to maintain the continuity of available leases.

4. The distributed object storage service quality assurance method based on lease according to claim 1 is characterized in that: Flow control is performed according to the parameters in the flow control lease.

5. A distributed object storage service quality assurance system based on lease, characterized in that: include: one or more processors; a storage device for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors implement the lease-based distributed object storage service quality assurance method according to any one of claims 1 to 4.

Citation Information

Patent Citations

  • Wireless multi-hop network bandwidth allocation method integrating time division and cache proportion

    CN104302008A

  • Flow limiting method and system

    CN110138756A

  • Lease enforcement in a distributed file system

    US20020152214A1

  • Scalable leases

    US20070156588A1

  • Method, controller, and system for service flow control in object-based storage system

    US20170220500A1