3. Configuration of Cluster Parallel Environments

INFO

All following operations are performed on the management node.
Before setting the relevant services, run the following commands:

sudo ./config_server.sh -pre

• In the podsys folder, generate a host.txt file. This file should contain the IP addresses of all nodes that can be successfully accessed via SSH. This is important for subsequent parallel environment configuration.
• This will ensure passwordless SSH login among the nexus users across all nodes.

3.1 Configuration of NFS

NFS Introduction

The Network File System (NFS) is a distributed file system protocol that allows a user on a client computer to access files and directories residing on a server in a network as if they were on their own local hard drive. It enables seamless sharing of files across multiple systems, facilitating a unified view of the filesystem across different nodes. NFS achieves this by exporting specific directories from the server, which can then be mounted and accessed remotely by clients. This protocol is widely used in large-scale environments such as data centers for its efficiency and scalability.

• Configuration commands:

Management Node

# Run on the management node
sudo ./config_server.sh -nfs [share directory]
# [share directory] indicates the folder that the management node needs to share.

Compute Node

# Run on the management node
sudo ./config_client.sh -nfs [serverIP] [share directory] [local directory]
# [serverIP] represents the IP address of the management node.
# [share directory] represents the shared folder of the management node.
# [local directory] represents the folder where the compute node mounts the nfs.

3.2 Configuration of NFSoRDMA

WARNING

NFSoRDMA requires IB network.

NFSoRDMA Introduction

NFS over Remote Direct Memory Access (NFSoRDMA) is an extension of the Network File System protocol that utilizes Remote Direct Memory Access technology to significantly enhance performance. By directly accessing memory in a remote system without involving the operating system, NFSoRDMA reduces latency and increases throughput for data-intensive applications, making it particularly suitable for high-performance computing environments where speed and efficiency are critical.

• Configuration commands:

Management Node

# Run on the management node
sudo ./config_server.sh -nfsordma [share directory]
# [share directory] indicates the folder that the management node needs to share.

Compute Node

# Run on the management node
sudo ./config_client.sh -IPoIB
sudo ./config_client.sh -nfsordma [serverIP] [share directory] [local directory]
# [serverIP] represents the IB IP address of the management node.
# [share directory] represents the shared folder of the management node.
# [local directory] represents the folder where the compute node mounts the nfs.

3.3 Configuration of NIS

TIP

NIS and OpenLDAP are both used for user management, and you need to choose one of them for configuration.

NIS Introduction

NIS: Network Information Service that was developed by Sun Microsystems. It allows system administrators to manage and distribute critical system information such as user names, passwords (in encrypted form), hostnames, IP addresses, network configurations, printer queues, and other related data across multiple computers in a local area network (LAN). NIS makes it easier for networked systems to maintain consistency in their configuration and settings. It is widely used in Unix and Unix-like operating systems.

• Configuration commands:

Management Node

# Run on the management node
sudo ./config_server.sh -nis [serverIP]
sudo /usr/lib/yp/ypinit –m
sudo make -C /var/yp

Compute Node

# Run on the management node
sudo ./config_client.sh -nis [nis server ip]
sudo yptest

Verification

# Create a new user on the management node
sudo adduser test
# Follow the prompts to enter a password, etc.
sudo make -C /var/yp
ll /home
su test

# Switch users on the compute node
su test
ll /home # Check if the user directory has been created

3.4 Configuration of OpenLDAP

TIP

NIS and OpenLDAP are both used for user management, and you need to choose one of them for configuration.

OpenLDAP Introduction

OpenLDAP is an open-source implementation of the Lightweight Directory Access Protocol (LDAP) that provides a directory service for storing and retrieving information about network entities. Developed by the OpenLDAP Project, it allows administrators to create centralized databases that store data like user accounts, group memberships, network resources, email addresses, and other critical information.

OpenLDAP supports a wide range of platforms including Unix-like systems, Windows, and various network environments. It is highly customizable and can be extended through plugins for additional functionality. Due to its flexibility and robustness, OpenLDAP is widely used in enterprise environments for managing user authentication, authorization, and access control.

• Configuration commands:

Management Node

# Run on the management node
sudo ./config_server.sh -ldap [serverIP] [ldap—password]

Compute Node

# Run on the management node
sudo ./config_client.sh -ldap [serverIP] [ldap—password]