Memory Fabric User Guide

Introduction

The Memory Fabric memory pool suite builds high-performance distributed memory resource pool based on memory medium and new network technologies, and provides services for applications through semantics such as BigMemory, MemPlog, and MemKV.

Memory Fabric uses and manages resource such as DRAM, persistent memory, and RAM drives, and provides microsecond-level latency based on new interconnection protocols (RDMA, XCL, and UBUS). By reducing the software latency and node CPU overhead, Memory Fabric improves the overall performance. (Pools of directly connect drives are not supported.)

Deployment

Memory Fabric supports multiple deployment modes for multi-scenario application integration based on the node types, resource distribution, and application requirements. Deployment modes include:

Co-deployed clients and servers when computing and Memory Fabric resources are on the same node. (Node1 and Node2 in Figure 1)
Independently deployed clients ands server when computing and Memory Fabric resources are on different nodes. (Node3 and Node4 in Figure 1)
Hybrid deployment of co-deployed clients and servers and independently deployed clients and servers.
Multiple clients and servers in one node. (Node2 in Figure 1)
Inter-process communication (IPC) for intra-node communication between clients and servers, and remote procedure call (RPC) for connection to remote servers.
Servers connected to the cluster management service for unified node numbering an resource information broadcasting.

Figure 1 Integrated deployment

Startup

To start Memory Fabric, you need to start the server and client. Start the server on the node first, perform the cluster registration and monitoring procedures, and initiate resource registration (the total capacity is obtained from the configuration file) and communication link establishment. After the initialization, multiple local clients can connect to the server and provide services for external systems.

Interfaces

Interface	Description
int BM_Init(char ockPath, int flags); int BM_InitWithExtAttr(char ockPath, int flags, void *attr); int BM_Exit(void);	Initializes Initializes with attributes Exits
int BM_CreateNSSec(const char poolName, const char ns, int attribute, const char secId); int BM_DestroyNSSec(const char poolName, const char ns, const char secId); int BM_QueryNSCache(QueryNsCache *buff, size_t buffLen);	Creates, destroys, and queries namespaces.
int BM_AllocSec(BmInfo bminfo, size_t size, int flags, BmAffinity affinity); int BM_FreeSec(BmInfo bminfo, int flags); int BM_BatchFreeSec(char bmIdArray[], int num, int flags, const char ns, const char secId);	Applies for and releases BigMemory objects
int BM_ExpandSec(BmInfo *bminfo, size_t size, int flags);	Expanses BigMemory capacity
int BM_MapSec(BmInfo bminfo, int prot, int flags, void ptr); int BM_Unmap(char bmId, void *ptr);	Maps BigMemory to continuous virtual space and unmaps
int BM_CopySec(BmInfo srcBM, size_t srcOffset, BmInfo dstBM, size_t dstOffset, size_t length, int flags);	Copies between BigMemory objects
int BM_SpecificObjSwapInSec(DiskFileDesc file, BmInfo bminfo, int flags, size_t offset, size_t length);	Swaps file content into the BigMemory object
int BM_ReadSec(BmInfo bminfo, size_t offset, size_t length, void buffer, int flags); int BM_WriteSec(BmInfo bminfo, size_t offset, size_t length, const void buffer, int flags);	Reads and write BigMemory
int BM_GatherSec(intptr_t inputInfo, BmOperatorCode operatorCode, int flags, BmInfo bminfo, size_t *bmLen);	Performs aggregation

Bug Catching

Buggy Content

Bug Description

Submit As Issue

It's a little complicated....

I'd like to ask someone.

Just a small problem.

I can fix it online!

Bug Type

Specifications and Common Mistakes

● Misspellings or punctuation mistakes;

● Incorrect links, empty cells, or wrong formats;

● Chinese characters in English context;

● Minor inconsistencies between the UI and descriptions;

● Low writing fluency that does not affect understanding;

● Incorrect version numbers, including software package names and version numbers on the UI.

Usability

● Incorrect or missing key steps;

● Missing prerequisites or precautions;

● Ambiguous figures, tables, or texts;

● Unclear logic, such as missing classifications, items, and steps.

Correctness

● Technical principles, function descriptions, or specifications inconsistent with those of the software;

● Incorrect schematic or architecture diagrams;

● Incorrect commands or command parameters;

● Incorrect code;

● Commands inconsistent with the functions;

● Wrong screenshots.

Risk Warnings

● Lack of risk warnings for operations that may damage the system or important data.

Content Compliance

● Contents that may violate applicable laws and regulations or geo-cultural context-sensitive words and expressions;

● Copyright infringement.

We may contact you through email in regard to the bug fixing solution. You will also be notified through email if you win any prize during the activity.

How satisfied are you with this document

Not satisfied at all

Very satisfied

By submitting the contents, you fully understand and agree to the terms of the openEuler Privacy Policy.

Submit

Click to create an issue. An issue template will be automatically generated based on your feedback.