- PRAGYA:
Master Node : 1 Processor : 2 * Intel(R) Xeon(R) Gold 6330 CPU @ 2.00GHz Cores : 56 Memory : 256 GiB DDR4 DRAM Storage : 7.64 TB SSD Compute Nodes : 3 Processor : 2 * AMD(R) EPYC 9654 96-Core Processor @ 2.40 GHz Cores : 192 Memory : 768 GiB DDR5 DRAM Storage : 1.92 TB SSD Operating System : AlmaLinux 8.9 64-bit Home Directory Storage : 1.80 TB Scratch Directory Storage : 3.50 TB - RUDRA:
Master Node : 1 Processor : 2 * Intel(R) Xeon(R) Gold 6330 CPU @ 2.00GHz Cores : 56 Memory : 256 GiB DDR4 DRAM GPU : 2 * NVIDIA Corporation GA102GL [A40] GPU Cores : 21504 CUDA Cores GPU Memory : 96 GB GDDR6 VRAM Storage : 12.36 TB SSD Operating System : Ubuntu 22.04.4 LTS 64-bit Home Directory Storage : 7 TB - Note:- All the specifications mentioned above are as per node.
- To begin with, the intending user needs to apply for the registration. Refer to (http://www.aus.ac.in/ict-data-centre/registration/) for the registration process.
- After verifying all the details, the account will be created.
- Then, user will receive an email in the official Email ID provided intimating the creation of the account along with a temporary password set by the system to the account.
- Login using the account credentials provided and do change the temporary password provided by the system to user’s own password. Once done, user is ready to use.
First login using the credentials and then the password can be changed using the passwd command: $ passwd.
- Accessing the cluster:
- The cluster can be accessed through Assam University, Silchar LAN IP address provided, which allows users to login.
- The master node (Entry node) may be accessed through ssh.
- The master node has a job scheduler (called SLURM) for managing jobs.
- For submitting jobs, always use a job script and sbatch command: $ sbatch job_script.sh. Refer to the “job_scipts” directory for available job scripts.
- Do not submit the jobs directly on the master node, as the system will automatically terminate such jobs. Sometimes, it may take more than a few minutes for the job to be run on the cluster, hence submit the job into the queue.
- Remote Access:
- Using ssh command:
- Open the terminal, connect using the ssh command: $ ssh [username]@[IP_address] (i.e; Assam University, Silchar LAN IP address provided).
- Then, it will prompt for a password, enter the user’s password and hence the ssh connection will be established.
- Using PuTTY application:
- Firstly, download the PuTTY application and install it.
- Once installed, launch the PuTTY application.
- The PuTTY Configuration dialog will appear.
- Locate the “Host Name (or IP Address)” input field.
- Enter the username along with IP address: [username]@[IP_address] (i.e; Assam University, Silchar LAN IP address provided).
- Enter the password when prompted and hence the connection will be established.
- Using ssh command:
- Using scp:
- To transfer files from Server to Local Machine, open the terminal and use the scp command:
scp -r [username]@[IP_address]:/[path_of_files_in_Server] /[path_in_Local_Machine] - To transfer files from Local Machine to Server, open the terminal and use the scp command:
scp -r /[path_of_files_in_Local_Machine] [username]@[IP_address]:/[path_in_Server]
- To transfer files from Server to Local Machine, open the terminal and use the scp command:
- Using WinSCP:
- Firstly, download the WinSCP application and install it.
- Once installed, launch the WinSCP application.
- The Login dialog will appear (Click ‘New Tab’, if the Login dialog does not appear).
- Click ‘New Site’ (if not selected).
- Locate the ‘Host name:’, ‘User name:’ & ‘Password:’ input field and enter the IP address, username & password respectively.
- Click ‘Login’ button and hence the connection will be established.
- Navigate to the locations of both the Local Machine and the Server on either sides.
- Finally transfer files using the ‘Drag-and-Drop’ feature.