HPC Support - faq

001

Anonymous (anonymous@undisclosed.example.com) — 2025-10-27T15:10:02+00:00

How do I run a job for longer than allowed? Sometimes it is necessary to run a job for longer than the defaults for a partition allow. In these cases we recommend you follow these steps: * Write your SBATCH job with the runtime possible for the partition you are submitting to (

002

Anonymous (anonymous@undisclosed.example.com) — 2025-08-06T09:56:51+00:00

How do I get an interactive shell on a compute node? Sometimes you may want to run some test code on a compute node, or compile something interactively on a compute node to ensure that your code works as intended before running the main job. You can request a short-term interactive shell session by running the following from a login node:

003

Anonymous (anonymous@undisclosed.example.com) — 2025-04-01T16:14:34+00:00

How Can I List My Current Scheduled Jobs? Also: “Why is my job not running yet?” You can get a list of jobs that you have submitted and which are waiting to run by using the squeue -t PD --me command: The PD option is a shorthand for jobs of status type

004

Anonymous (anonymous@undisclosed.example.com) — 2025-03-11T13:38:26+00:00

How Can I See Which of My Jobs Have Failed Recently? You can retrieve a report on all jobs which have failed within the last 30 days using the following variation of the sacct command: You can adjust the time window using the --starttime and --endtime

005

Anonymous (anonymous@undisclosed.example.com) — 2025-04-16T08:34:53+00:00

How Can I See Which Of My Jobs Are Running? It is a common task to see which of your jobs are currently running. You can achieve this with squeue -t R --me: Note: Where the state type R indicates the job is RUNNING. ---------- Back to FAQ

006

Anonymous (anonymous@undisclosed.example.com) — 2025-03-18T14:43:30+00:00

How can I report on all home directory use? If you are a system administrator, you can get a CSV listing of all home directory use, including status of the owners of home directories with the following script: You will need to set the following variables:

007

Anonymous (anonymous@undisclosed.example.com) — 2025-03-19T15:34:55+00:00

How do I set up Github authentication on the HPC? Lots of our users need the ability to clone code from , so that they can build software themselves, for their own use. Most repositories are in the process of moving over to Git+SSH protocol, rather than the legacy HTTPS access method. In order to commit (or even clone, in many circumstances) you will need an SSH key setting up on our HPC facilities in order to authenticate with Github.

008

Anonymous (anonymous@undisclosed.example.com) — 2025-04-09T13:02:15+00:00

Make A Script Executable Normally, to run a script on Linux you would have to call it with the name of the application, for example with a Python script myscript.py you would type something like: ... or, if it was a shell script named myscript.sh

009

Anonymous (anonymous@undisclosed.example.com) — 2025-04-09T13:06:19+00:00

Set Permissions On Files Or Directories * * Files Permissions on Linux are applied to files by three main groups of three attributes: * r indicates the ability to read the contents of the file * w indicates the ability to write to or modify the file

010

Anonymous (anonymous@undisclosed.example.com) — 2025-04-09T12:39:05+00:00

Viewing Files & Comparing File Contents * - View text file, with scrolling and searching * - View top n lines of a text file * - View last n lines of a text file * - Compare contents of two files Viewing Files Let's assume we have the file

011

Anonymous (anonymous@undisclosed.example.com) — 2025-06-25T15:01:54+00:00

How can I see what resources are on each node/partition? Option 1 View the HPC resources type table, this gives you detailed hardware specifications of each physical server node type: * ---------- Option 2 View the partition definitions, this shows what default values are available in each of the Slurm resource partitions.

012

Anonymous (anonymous@undisclosed.example.com) — 2025-04-01T16:03:11+00:00

Help! I cannot access RDW Losing access to RDW is a common issue that users of the HPC facilities raise. Normally this isn't really a problem with the service, but a peculiar feature of how the Campus Research Data Warehouse is implemented - remember; RDW is

013

Anonymous (anonymous@undisclosed.example.com) — 2025-04-28T08:21:20+00:00

How much does RDW storage cost? Research Data Warehouse (RDW) provides a secure place for research data, which can be accessed from Rocket. * Find out more at RDW filestore description. * Check current costs and terms at RDW filestore service definition. Please be aware that RDW is not part of the HPC facility and the service is not supported by the Research Software Engineering / Research Computing service.

014

Anonymous (anonymous@undisclosed.example.com) — 2025-04-28T08:15:52+00:00

Why is my home directory over quota? Information relates to Rocket HPC, may be updated after Comet comes online Home directory quota is 40GB for everyone - that should be plenty of space for installing software. Working data should not live in HOME

015

Anonymous (anonymous@undisclosed.example.com) — 2025-05-02T09:06:52+00:00

Individual logs for task array and srun jobs There are a myriad of ways of capturing output from running Slurm tasks. Normally all output, regardless of number of parallel tasks is captured by the main Slurm output file. Below are three additional options to alter your Slurm job script, to capture

016

Anonymous (anonymous@undisclosed.example.com) — 2025-06-05T17:52:26+00:00

Centrally installed Software The full software list may be found at HPC Documentation wiki -> Advanced -> Software -> Software list June 2025: Please note that this software will be installed over a period of time as the new Comet HPC service is brought up. Essential and widely used software will be prioritised. Not all software will be available on Day 1 of Comet availability.

017

Anonymous (anonymous@undisclosed.example.com) — 2025-12-17T12:32:26+00:00

New Software Installation Requests Please check the Comet Software List, which shows all software planned for installation by the HPC Replacement Project Team. Software installation is taking place alongside other setup by our supplier so not all software will be available on Day 1. For software not on this list, requests can be made via

18

Anonymous (anonymous@undisclosed.example.com) — 2025-07-04T14:20:48+00:00

What hardware is available on Comet? Comet has a mix of CPU and GPU nodes, all with x86_64 architecture. The HPC Portal has a list of resources and links to a calculator tool to help you work out the amount of resources you will need, and how much you might pay for enhanced access. Remember that the free tier of Comet will still provide more resource than Rocket.

019

Anonymous (anonymous@undisclosed.example.com) — 2025-12-11T11:24:58+00:00

Gathering Troubleshooting Information For troubleshooting either on your own or with assistance from others, you will want to know “What, Where, When, and How” an error occurred. It's a good idea to copy and paste from the terminal, to capture the exact time, what you were doing, the hostname and current directory.

020

Anonymous (anonymous@undisclosed.example.com) — 2025-11-24T17:04:00+00:00

Research Data Warehouse Research Data Warehouse (RDW) is resilient storage on campus, intended for long term storage of research data. RDW is mounted at /rdw on the Rocket and Comet login nodes. * Read our Data and HPC filesystems for more detail on how RDW is used and complements our other HPC filesystems

021

Anonymous (anonymous@undisclosed.example.com) — 2025-07-16T14:43:39+00:00

When Should I Use the Low Latency Nodes? If you are using MPI jobs then you may immediately think that the low latency nodes are what you should be using. Instinctively this would appear to be true, however the truth is a little more nuanced. If your MPI job fits entirely within the resource constraints of a single node, then you will get

022

Anonymous (anonymous@undisclosed.example.com) — 2025-10-09T10:21:04+00:00

Where do I put my HPC data? For general advice about research data storage (non-HPC specific) within the University, contact the Research Data Mangement team: rdm@ncl.ac.uk Find out about our HPC filesystems on the Data And HPC Filesystems page. ---------- Back to FAQ index

025

Anonymous (anonymous@undisclosed.example.com) — 2025-10-09T10:33:13+00:00

Can my collaborator from another institution use HPC at Newcastle? Only users with a long term Newcastle University IT account (more than 4 weeks) can be granted access to our HPC. To find out more about registering a new Newcastle University IT account, please check the following links:

026

Anonymous (anonymous@undisclosed.example.com) — 2025-10-09T10:22:52+00:00

How can I move my data to where I need it? Moving data to, from and on our HPC is covered within our Getting Started section, on the Data Transfer page. ---------- Back to FAQ

027

Anonymous (anonymous@undisclosed.example.com) — 2026-02-23T11:12:53+00:00

How can I check what resources my finished jobs used? Knowing what resources your jobs actually used is key to requesting the optimal resources for future jobs. A quick and easy option is seff outputs You can check individual job steps with the syntax

028

Anonymous (anonymous@undisclosed.example.com) — 2025-10-09T10:29:35+00:00

How can I get an email when my job completes? To receive an email when your job completes, include the following lines in your sbatch script. Replace with your own NU email address. ---------- Back to FAQ

029

Anonymous (anonymous@undisclosed.example.com) — 2025-08-28T12:59:22+00:00

Open On Demand: Home Directory not found Open on Demand is a graphical web interface that allows you to run grahical and interactive applications on Comet HPC. Users who have never logged in to Comet via the command line will not have had a home directory created yet.

030

Anonymous (anonymous@undisclosed.example.com) — 2025-09-03T11:11:17+00:00

Check your home directory quota and usage The quota -s command can be used to check your quota. space and files show what is actually used, while quota, limit and grace fields refer to the limits that have been set by the system administrator. For example, user abc123 below:

031

Anonymous (anonymous@undisclosed.example.com) — 2025-10-30T11:31:14+00:00

Why does my MPI job complain about connection errors or addresses already in use? Depending on the number of MPI processes you attempt to start in your Slurm jobs, you may see errors such as: Or Or words to the effect of: This occurs because MPI processes open network sockets to talk to each other. Depending on the MPI communications mechanism your application is configured to use,

032

Anonymous (anonymous@undisclosed.example.com) — 2025-11-03T12:29:16+00:00

Account Code & QOS Policy Errors QOS Policy Errors If you see this error: This indicates that you did not submit your job with the --account= parameter, and Slurm has attempted to use your default account code. Every Slurm user on Comet has the

033

Anonymous (anonymous@undisclosed.example.com) — 2026-02-12T09:50:49+00:00

How can I tell which account codes I can use and which partitions I can submit to? The account codes you can use when submitting a job to Slurm are directly related to the HPC Projects you are a member of (see My HPC Projects). * All HPC Projects that are: *

034

Anonymous (anonymous@undisclosed.example.com) — 2025-11-28T11:41:48+00:00

What Is The HPC Driving Test? As is common with many other HPC facilities, the University has implemented a short interactive test for anyone needing access to HPC resources. * All HPC users must pass (100% correct) the HPC driving test * You must pass this test

035

Anonymous (anonymous@undisclosed.example.com) — 2025-12-05T09:02:59+00:00

I enter the correct password but cannot login! If you are attempting to SSH to the Comet HPC facility using the connection details we provide, and Comet refuses to allow you to log on, and all of the following are true: * You have passed the test, as shown in the

36

Anonymous (anonymous@undisclosed.example.com) — 2026-01-05T17:01:34+00:00

Why do I see 'Connection Closed' when I try to log in to Comet? 'Connection Closed' is the message you receive when you try to ssh to a server and your access is denied. This can happen for a variety of reasons: * You have only just been added to your first Comet project - it takes 24 hours for project membership to be processed automatically and membership of at least 1 project is needed before Comet will accept your login

036

Anonymous (anonymous@undisclosed.example.com) — 2026-01-21T11:36:04+00:00

How to get rid of a red box error in front of your Open OnDemand Application Before connecting to your session in Open OnDemand, you can adjust sliders for compression and image quality. There is a tradeoff here, higher image quality requires more bandwidth, while higher compression uses more CPU but reduces bandwidth usage. The best settings for you will vary depending on your network connection. However, setting Compression to

037

Anonymous (anonymous@undisclosed.example.com) — 2026-01-22T17:15:18+00:00

Why is my file operation going so slowly? Most people don't realise that HPCs have a variety of filesystems optimised for different tasks. This blog post highlights the differences that can occur: * Another factor that's often missed is that the number of files in a directory can have a dramatic effect on some operations on the Lustre (/nobackup) filesystem. If you have large numbers (thousands) of very small files, it's best to split them into multip…

038

Anonymous (anonymous@undisclosed.example.com) — 2026-03-11T09:18:05+00:00

How do I extend my LD_LIBRARY_PATH for R/R-Studio? When launching R-Studio from our Open OnDemand service you don't have the ability to preload any custom modules or any paths to self-installed code. Once R-Studio is running it's not feasible to amend the LD_LIBRARY_PATH

039

Anonymous (anonymous@undisclosed.example.com) — 2026-04-09T14:36:31+00:00

How Do I Install More Python Modules? All users of Comet may install their own, additional Python modules without the need to contact RSE HPC support. Please see our guide below: * See: Python Modules Installation ---------- Back to FAQ index

040

Anonymous (anonymous@undisclosed.example.com) — 2026-04-09T15:20:03+00:00

How Do I Install More R Libraries? All users of Comet may install their own, additional R libraries without the need to contact RSE HPC support. Please see our guide below: * See: R Library Installation ---------- Back to FAQ index

files_attrs_were_not_transferred

Anonymous (anonymous@undisclosed.example.com) — 2025-10-08T15:22:14+00:00

files/attrs were not transferred files/attrs were not transferred: This error may be returned because RDW doesn't 'know' about Rocket's groups. * Applying group permissions from Rocket will fail because RDW has 'trumped' our local permissions and imposed its own permissions.

index

Anonymous (anonymous@undisclosed.example.com) — 2026-04-09T14:34:35+00:00

Frequently Asked Questions Creating and Managing HPC Projects * What is the HPC Driving Test? Data Storage and Transfer * Where do I put my HPC data? * How can I use Research Data Warehouse (RDW) with HPC? * How can I move my data to where I need it? Filesystems * Help! I cannot access RDW * How much does RDW cost? * Why is my gzip / tar / cp operation going SO SLOWLY? Git & Github for managing code and scripts * How do I set up Github.com authentication on the HPC? Ope…

long_transfers_being_halted_by_permissions_errors

Anonymous (anonymous@undisclosed.example.com) — 2025-10-08T15:28:41+00:00

Long transfers being halted by permissions errors This usually means the ‘kerberos ticket’ for your user has expired. Kerberos tickets allow the system to know what your user is allowed to do, for security, they automatically expire at a set time after you log in. Most of the time these tickets are automatically renewed while you’re working, but they can expire during long copy commands.

on_campus

Anonymous (anonymous@undisclosed.example.com) — 2025-10-08T14:59:32+00:00

On Campus Systems connected to the campus network include: * Comet HPC * Rocket HPC * Windows Virtual Desktop * Workstations in campus buildings, connected via eithernet cable * Laptops connected to campus WiFi (NOT guest WiFi)

rdw_setup

Anonymous (anonymous@undisclosed.example.com) — 2025-11-24T17:20:19+00:00

Setting up a new RDW share for a research project RDW is the best place to store research data, it is accessible as a mount on both Rocket and Comet HPCs and also from campus Linux and Windows computers. To request a new RDW share “Request Something

rdw_transfer_how-to

Anonymous (anonymous@undisclosed.example.com) — 2026-04-29T14:47:19+00:00

File Transfers and RDW Monitoring RDW shares RDW is file storage for research, hosted on our campus datacentre. It is possible to view the files from both Windows and Linux, permissions on the files are set using your campus login so it’s essential that you are logged in as the correct campus user in order to see your files.

windows_file_explorer

Anonymous (anonymous@undisclosed.example.com) — 2026-03-12T15:51:50+00:00

Viewing RDW On Windows Windows File Explorer is NOT suitable for copying data to or from Rocket. See File Transfer and RDW for the correct method. Windows File Explorer or Powershell can be a handy way to check that data exists on RDW. * Browse your files under the path