CN111352578A - Memory borrowing strategy between brand new servers - Google Patents
Memory borrowing strategy between brand new servers Download PDFInfo
- Publication number
- CN111352578A CN111352578A CN201811580976.5A CN201811580976A CN111352578A CN 111352578 A CN111352578 A CN 111352578A CN 201811580976 A CN201811580976 A CN 201811580976A CN 111352578 A CN111352578 A CN 111352578A
- Authority
- CN
- China
- Prior art keywords
- memory
- servers
- borrowing
- rdma
- brand new
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
- G06F3/067—Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F13/00—Interconnection of, or transfer of information or other signals between, memories, input/output devices or central processing units
- G06F13/14—Handling requests for interconnection or transfer
- G06F13/20—Handling requests for interconnection or transfer for access to input/output bus
- G06F13/28—Handling requests for interconnection or transfer for access to input/output bus using burst mode transfer, e.g. direct memory access DMA, cycle steal
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0629—Configuration or reconfiguration of storage systems
- G06F3/0631—Configuration or reconfiguration of storage systems by allocating resources to storage systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0662—Virtualisation aspects
- G06F3/0664—Virtualisation aspects at device level, e.g. emulation of a storage device or system
Abstract
The invention relates to the technical field of information, in particular to a memory borrowing strategy between brand new servers, which mainly comprises the following steps: s1: connecting the servers by adopting an infiniband network card, and enabling the transmission speed between the servers to be more than or equal to 10 GB/S; s2: the memory borrowing between the servers adopts RDMA as a transmission protocol, because the RDMA can directly read the memory data of the remote server and does not pass through the CPU of the server of the other side; s3: simulating a local disk by the remote memory through the ISER; s4: and then setting the simulated local disk as swap space. The invention provides a memory borrowing strategy between brand new servers, which adopts an RDMA mode to use a remote memory as a swap space of a local machine, improves the transmission speed of the swap space and achieves the purpose of memory borrowing between servers.
Description
Technical Field
The invention relates to the technical field of information, in particular to a memory borrowing strategy between brand new servers.
Background
In a server cluster, some server memories are insufficient, and some server memories are low in use rate. If the server memory with low utilization rate can be efficiently borrowed and used as the server with insufficient memory, the resource utilization rate of the server cluster can be improved.
Disclosure of Invention
In order to solve the problems in the background technology, the invention provides a memory borrowing strategy between brand new servers, remote memories are used as swap spaces of the servers in an RDMA mode, the transmission speed of the swap spaces is improved, and the purpose of memory borrowing between the servers is achieved.
The technical scheme for solving the problems is as follows: a memory borrowing strategy between brand new servers is characterized by comprising the following steps:
s1: providing at least two servers, wherein the at least two servers are connected by adopting an infiniband network card, and the transmission speed between the at least two servers is more than or equal to 10 GB/S;
s2: memory borrowing between at least two servers adopts RDMA as a transmission protocol; because RDMA can directly read the memory data of the remote server without going through the CPU of the opposite server;
s3: simulating a local disk by the remote memory through the ISER;
s4: and then setting the simulated local disk as swap space.
The invention has the advantages that:
according to the memory borrowing strategy between the brand-new servers, the remote memory is used as the swap space of the local machine in an RDMA (remote direct memory access) mode, the transmission speed of the swap space is improved, the purpose of memory borrowing between the servers is achieved, and the overall memory utilization rate of a server cluster is improved.
Drawings
Fig. 1 is a full physical configuration diagram in the embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings of the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention. Thus, the following detailed description of the embodiments of the present invention, presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
A memory borrowing strategy between brand new servers comprises the following steps:
s1: providing at least two servers, wherein the at least two servers are connected by adopting an infiniband network card, and the transmission speed between the at least two servers is more than or equal to 10 GB/S;
s2: memory borrowing between at least two servers adopts RDMA as a transmission protocol; because RDMA can directly read the memory data of the remote server without going through the CPU of the opposite server;
s3: simulating a local disk by the remote memory through the ISER;
s4: and then setting the simulated local disk as swap space.
The following detailed description of the memory borrowing policy between the brand-new servers of the present invention is made with reference to fig. 1:
as shown in fig. 1, the memory of the physical machine 1 is 96GB, and the memory of the physical machine 2 is 64GB, and infiniband network cards are used for connection therebetween. The hardware requirement of the physical machine 1 borrowing the memory of the physical machine 2 is as follows: by using the memories of other physical machines, the most important index is the transmission speed, the transmission speed can reach 10GB/S based on the infiniband network card, and the reading speed of a common SSD hard disk is about 500M/S. Therefore, infiniband connection between physical machines is a prerequisite for borrowing the memory. The following describes the procedure of memory borrowing.
In the first step, the memory block of the physical machine 2 is simulated as the local disk of the physical machine 1. This functionality can be implemented by ISER, which is an RDMA-based extension of the ISCSI protocol (iSCSI utilizes TCP/IP ports 860 and 3260 as a conduit for communication.
And secondly, setting the simulated local disk as a swap space. Because the Infiniband-based RDMA transmission speed can reach 10GB/S, which is equivalent to 20 times of the transmission speed of a common solid state disk.
The remote memory can be well borrowed by simulating the remote memory into a swap space. Meanwhile, the bottom layer transmission adopts the RDMA based on infiniband, so that the transmission speed can be well ensured.
According to the memory borrowing strategy between the brand-new servers, the remote memory is used as the swap space of the local machine in an RDMA (remote direct memory access) mode, the transmission speed of the swap space is improved, the purpose of memory borrowing between the servers is achieved, and the overall memory utilization rate of a server cluster is improved.
The above description is only an embodiment of the present invention, and not intended to limit the scope of the present invention, and all equivalent structures or equivalent flow transformations made by using the contents of the specification and the drawings, or applied directly or indirectly to other related systems, are included in the scope of the present invention.
Claims (1)
1. A memory borrowing strategy between brand new servers is characterized by comprising the following steps:
s1: providing at least two servers, wherein the at least two servers are connected by adopting an infiniband network card, and the transmission speed between the at least two servers is more than or equal to 10 GB/S;
s2: memory borrowing between at least two servers adopts RDMA as a transmission protocol;
s3: simulating a local disk by the remote memory through the ISER;
s4: the emulated local disk is set to swap space.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811580976.5A CN111352578A (en) | 2018-12-24 | 2018-12-24 | Memory borrowing strategy between brand new servers |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811580976.5A CN111352578A (en) | 2018-12-24 | 2018-12-24 | Memory borrowing strategy between brand new servers |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111352578A true CN111352578A (en) | 2020-06-30 |
Family
ID=71195513
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811580976.5A Pending CN111352578A (en) | 2018-12-24 | 2018-12-24 | Memory borrowing strategy between brand new servers |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111352578A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102404212A (en) * | 2011-11-17 | 2012-04-04 | 曙光信息产业(北京)有限公司 | Cross-platform RDMA (Remote Direct Memory Access) communication method based on InfiniBand |
WO2016101288A1 (en) * | 2014-12-27 | 2016-06-30 | 华为技术有限公司 | Remote direct memory accessmethod, device and system |
CN106844048A (en) * | 2017-01-13 | 2017-06-13 | 上海交通大学 | Distributed shared memory method and system based on ardware feature |
US20170277655A1 (en) * | 2016-03-25 | 2017-09-28 | Microsoft Technology Licensing, Llc | Memory sharing for working data using rdma |
-
2018
- 2018-12-24 CN CN201811580976.5A patent/CN111352578A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102404212A (en) * | 2011-11-17 | 2012-04-04 | 曙光信息产业(北京)有限公司 | Cross-platform RDMA (Remote Direct Memory Access) communication method based on InfiniBand |
WO2016101288A1 (en) * | 2014-12-27 | 2016-06-30 | 华为技术有限公司 | Remote direct memory accessmethod, device and system |
US20170277655A1 (en) * | 2016-03-25 | 2017-09-28 | Microsoft Technology Licensing, Llc | Memory sharing for working data using rdma |
CN106844048A (en) * | 2017-01-13 | 2017-06-13 | 上海交通大学 | Distributed shared memory method and system based on ardware feature |
Non-Patent Citations (1)
Title |
---|
SHUANG LIANG ET AL: "Swapping to Remote Memory over InfiniBand: An Approach using a High Performance Network Block Device" * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104380278B (en) | Equipment, system and method for client-side management session continuity | |
EP3234797B1 (en) | System on a chip comprising reconfigurable resources for multiple compute sub-systems | |
US20080022120A1 (en) | System, Method and Computer Program Product for Secure Access Control to a Storage Device | |
CN104603739A (en) | Block-level access to parallel storage | |
CN109196829A (en) | remote memory operation | |
CN103257941B (en) | Multi-protocol storage controller and system | |
CN105867843A (en) | Data transmission method and device | |
CN107124469B (en) | Cluster node communication method and system | |
CN103176833A (en) | Data transmission method, data receiving method and system based on virtual machine | |
CN103703441A (en) | Hard disk and data processing method | |
CN104639469A (en) | Computing and storing cluster system based on PCIE (Peripheral Component Interconnect Express) interconnection | |
EP3995961A1 (en) | Storage system and operating method thereof | |
CN102413183B (en) | Cloud intelligence switch and processing method and system thereof | |
US20110246600A1 (en) | Memory sharing apparatus | |
WO2015180650A1 (en) | Method and device for generating key value in storage area network | |
CN102917005A (en) | Method and device supporting massive memory access to transactions | |
CN106648838B (en) | Resource pool management configuration method and device | |
CN104077374B (en) | A kind of method and device for realizing the storage of IP disk files | |
CN108282516B (en) | Distributed storage cluster load balancing method and device based on iSCSI | |
CN103067500B (en) | Web browsing method based on cloud terminal | |
CN108234551A (en) | A kind of data processing method and device | |
EP2309701A1 (en) | Data storage device | |
CN108306930A (en) | Remote desktop system and image data sharing method | |
CN100353330C (en) | Disk mirroring method based on IP network | |
CN110022341A (en) | A kind of data transmission method and relevant device |
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 |