= General =

There are a few options researchers can take advantage of when storing their research data.

== Data Classification ==
Please review the different data classifications that are outlined by the [https://ucalgary.ca/legal-services/sites/default/files/teams/1/Standards-Legal-Information-Security-Classification-Standard.pdf Information Security Classification Standard]. There are 4 levels of data classification which are summarized in the table below.

{| class="wikitable"
! Level
! Description
! Example
|-
| Level 1
| Public
|
* Reference data sets
* Published research data
|-
| Level 2
| Internal
|
* Internal memos
* Unpublished research data
* Anonymized or de-identified human subject data
* Library transactions and journals
|-
| Level 3
| Confidential
|
* Faculty/staff employment applications, personnel files, contact information
* Donor or prospective donor information
* Contracts
* Intellectual property
|-
| Level 4
| Restricted
|
* Patient identifiable health information
* identifiable human subject research data
* information subject to special government requirements
|}

When selecting a storage option, you must use one that meets or exceeds the rated security classification.

* See also the Collaboration, storage and file shares article in Service Now:
: https://ucalgary.service-now.com/it?id=it_catalog_by_category&sys_id=4dbb82ee13661200c524fc04e144b044

== Research Data Management ==
We recommend you follow good Research Data Management practices and ensure you have a DMP (Data Management Plan) created to guide your data's lifecycle. DMP Assistant has been created specifically for Canadian scholars and aims to meet any and all Tri-Agency requirements. See: https://assistant.portagenetwork.ca/

Your DMP can help us support the FAIR (findable, accessible, interoperable and reusable) principles for data management.

Please consider contacting Libraries and Cultural Resources for assistance. For guidance on general data management and developing a DMP, consult https://library.ucalgary.ca/guides/researchdatamanagement or contact research.data@ucalgary.ca.

For support using PRISM Dataverse, UofC's institutional data repository, contact digitize@ucalgary.ca.

If you need to share and preserve your large post-publication data set for a mandated period of time, please visit https://www.frdr-dfdr.ca/repo/ in order to learn more about the national Federated Research Data Repository.

FRDR aligns with Tri-Agency Principles as a platform for Preservation, Retention and Sharing of research data. see: [http://www.science.gc.ca/eic/site/063.nsf/eng/h_83F7624E.html Tri-Agency Statement of Principles on Digital Data Management]

= University of Calgary RCS storage services =

== Secure Compute Data Storage (SCDS) ==
Secure Computing Data Storage (SCDS) is a service provided by Research Computing Services that allows researchers to store restricted and confidential data. Collaboration with Level 4 data stored in SCDS is possible using ShareFile, a secure file sharing and collaboration tool by Citrix.

{| class="wikitable"
! Capacity
| 10 GB or more
|-
! Classification
| Level 4
|-
! Learn More
| Visit [https://it.ucalgary.ca/secure-computing-platform The SCDS Website]
|-
! Request Access
| Visit [https://ucalgary.service-now.com/kb_view.do?sysparm_article=KB0030163 ServiceNow to request access]
|}

== ResearchFS ==
ResearchFS is a UofC hosted SMB/CIFS storage solution funded and operated by RCS. It is available by request to faculty and staff with active research data.
{| class="wikitable"
! Capacity
| 1TB with quota increases available on request.
|-
! Classification
| Level 1 - 2
|-
! Request Access
| Visit [https://ucalgary.service-now.com/it?id=sc_cat_item&sys_id=fe66b3a7db297300897e4b8b0b96199d ServiceNow to request access]
|}
=== Service Description ===
You may use ResearchFS to store your active research data files. ResearchFS is intended to be used as a research group or project share. ResearchFS is available on campus or off campus using the IT supported VPN client. Information on how to download and install the VPN client can be found here: https://ucalgary.service-now.com/it?id=kb_article&sys_id=880e71071381ae006f3afbb2e144b05c (IT account login may be required).
All ResearchFS users must have a UofC IT account.

=== Data recovery ===
ResearchFS does daily snapshots at a bit past midnight, which it keeps for 30 days. You should be able to recover a deleted file for up to 30 days, if it was in your share overnight. If you create a file and delete it during a day, no snapshot will be available for you to recover. ResearchFS presents backups using the windows OS 'previous versions' functionality. If you are not familiar with using this, or if you are on a Linux or MacOS device, you can request a restore, with Service Now.

For backup, we replicate changes to a distant data center every hour. The storage hardware which hosts your data is located in the basement of the Math Sciences building and our backup is in the HRIC building, so in case of an on campus disaster, your data should be safe.

=== Support for ResearchFS ===
If you have questions, please contact the IT Support Centre.
: Mon – Fri: 8:30 am – 5:00 pm; Sat, Sun & holidays: 10:00 am – 2:00 pm.
: Live Chat: ucalgary.ca/it
: Email: itsupport@ucalgary.ca
: Phone: 403.220.5555
: In person: 773 Math Science

= University of Calgary IT storage services =

== OneDrive for Business ==
OneDrive for Business is a storage solution provided by Microsoft and is available to all faculty and staff.
{| class="wikitable"
! Capacity
| 5 TB with quota increases [https://ucalgary.service-now.com/it?id=sc_cat_item&sys_id=438e6d8313896a0053f2d7b2e144b0b9 available on request].
|-
! Classification
| Level 1 - 4
|}

You may use OneDrive for Business to store your personal and work related files.
Files stored within OneDrive are by default private only to you but has the option to allow sharing and collaboration with others.
OneDrive for Business cannot be used as a department or project share space.
There is no group/lab offering with OneDrive.

While OneDrive provides a secure/compliant location from an IT Security stand point,
it’s not the most adequate location for data the PI is accountable for 5 years upon completion of the study.
This is not a security issue, but a data management issue.

For example, if a study was using a personal OneDrive of one of the researchers to store all the records,
and the researcher was to leave the university, this OneDrive would be gone in 30 days.

MS has an automation capability for their O365 products.
If you have a windows OS machine, you can use the automation product ‘Flow’ to copy a file to a local file system when a new file is created on OneDrive.

To back up data residing on ARC to your personal OneDrive allocation please see: [[How to transfer data#rclone: rsync for cloud storage]]

OneDrive requires Multi-Factor Authentication (MFA) enabled on your University of Calgary IT account.

More information can be located in the following article https://ucalgary.service-now.com/kb_view.do?sysparm_article=KB0032351

UofC OneDrive data is reportedly hosted in Canada (Markham Ont).

===Support for OneDrive===
If you have questions, please contact the [https://ucalgary.service-now.com/it UService Support Centre].
: Email:it@ucalgary.ca Phone:403.210.9300 1.888.342.3802 Mon - Fri: 8:00 a.m. to noon and 1:00 p.m. to 4:30 p.m. (closed over the lunch hour) Walk-in service Math Sciences 7th floor, Room 773 Tues - Thurs: 1:00 p.m. to 4:30 p.m.

===Data recovery===

===Other Resources===
For more information on OneDrive for Business:
* Operating Level of Agreement KB0032404 (https://ucalgary.service-now.com/it?id=kb_article&sys_id=7f57bddcdb56a3047cab5068dc9619b6)
*OneDrive for Business Getting Started KB0032351 (https://ucalgary.service-now.com/it?id=kb_article&sys_id=60994170db2da7487cab5068dc961900)
*If you are above 90% of your OneDrive quota, you can request an increase here: ( https://ucalgary.service-now.com/it?id=sc_cat_item&sys_id=438e6d8313896a0053f2d7b2e144b0b9) PLEASE NOTE: Microsoft will only increase an allocation while the Cloud Storage is more than 90% full. Please log into your O365 cloud account to review before making your request.

Any questions regarding if data hosted on OneDrive is subject to US jurisdiction discovery or access should be directed to:
*https://cumming.ucalgary.ca/research-institutes/csm-research-services/legal-research-services (CSM researchers.)
*https://research.ucalgary.ca/contact/research-services (Not CSM Researchers)
*https://www.ucalgary.ca/legalservices/ (for teaching/learning – non research enquiries that make their way to you)

==Office365 SharePoint for research groups==

To be determined....

Researchers will be able to request an Office 365 SharePoint site for a group at some point in the future
which could be considered a group cloud sharing platform.

* The official service page:
: https://ucalgary.service-now.com/it?id=sc_cat_item&sys_id=b55f2f72132f5240b5b4df82e144b085

= Digital Research Alliance of Canada storage services =

== Storage on the Alliance HPC clusters ==

* Alliance Wiki article "Storage and file management":
: https://docs.alliancecan.ca/wiki/Storage_and_file_management

==Personal storage options==
For personal or level 1 data, you may use an external solution from the Alliance.
One has to have an Alliance account to use the service.
This is similar to DropBox or Google drive functionality.

*'''The Alliance NextCloud''':
:https://nextcloud.computecanada.ca
: 100 GB of storage that can be shared between your computers.
: Alliance documentation: https://docs.alliancecan.ca/wiki/Nextcloud

= Commercial Cloud Based Storage Options =

== AWS ==

Provided by '''Amazon Web Services, Inc.'''.

* Pricing calculator: https://calculator.aws

AWS provides very many different kinds of services, including '''storage services'''.

These options can be a solution for your research needs, but
* it '''can be expensive''', depending on your needs and the amount of data;
* '''pricing schema is complex''' and can be confusing for new users;
* the '''number of options can be overwhelming'''. A lot of it is designed to provide pricing flexibility, not to increase functionality.

The key points:
* '''Uploading data''' to AWS is '''free'''.
* '''Storing data''' on AWS storage is a '''paid service'''.
* '''Downloading data''' from AWS storage to your computer is a '''paid service'''.

= ARC Cluster Storage =

ARC storage is used to '''support workflows on the ARC computing cluster'''. The expectation is that storage on ARC will only be used for active and upcoming computational projects.
It is not suitable for long-term or archival storage as it is not backed-up and is not guaranteed to be available for the time periods that are typical of archiving.
ARC is a research cluster, which means it has high performance but can be stopped for required maintenance when needed.
Thus, ARC cannot be relied on for any kind of service that requires constant availability.
Which means, in turn that ARC's storage cannot and '''should not be used as a main storage facility for research data'''.
The '''master copy''' of research data should be '''stored elsewhere''' and only part of that data are expected to be copied to ARC for computational analysis.

== Home Directories ==
Every user account on ARC has a static 500GB allocation of storage and a maximum of 1.5 million files (including directories). This cannot be increased or decreased. Home directory storage is connected via a network file system to the rest of the cluster and supports fast data transfer to memory on compute nodes. This also means that basic file system commands (like <code>ls</code>, <code>find</code>, and <code>du</code>) take longer to run as the number of files in your home directory increases. In particular, we strongly encourage users to stay under 100000 files if it is at all possible. This can be achieved by combining smaller data files into single larger files, using structured data formats rather than large number of text files, or combining collections of files that will be used together into archives (tar, dar, etc). Since top level permissions on home directories are set to prevent other users from reading or executing, home directories are not suitable for sharing data directly with colleagues working on ARC. A Research Group Allocation is a more appropriate place for storing shared data or very large data sets that will be used as part of active computational projects.

== Research Group Allocations (<code>/work</code> and <code>/bulk</code>) ==
The principal investigator (PI) for a research group may request an extended shared allocation for the research group by contacting support@hpc.ucalgary.ca with answers to the following questions (please copy the full text of the questions into your email and write answers under it):

* How much storage is requested and why is that the amount that you need?
A rationale for a request can be a formal data management plan or something more informal like a rough estimate to the primary dataset used for a project and a rough estimate to the size of outputs expected from your computations that are planned to run on ARC over the '''next year.'''

* What is the requested '''allocation name'''? (typically something like <PI name>_lab, <code>smith_lab</code>, for example)
* What is the '''data classification''' using the University of Calgary data security classification system?
* Which user or users would be the '''owner''' of the allocation? (Full Name and UCalgary Email address, typically the requesting PI but there may be co-PIs)
* Which members of the allocation should be able to '''request access''' for new users? (Full Name and UCalgary Email address for active ARC users)
* What is the '''faculty''' of the owner or owners?
* Please provide a short description of the lab.
* Please provide a brief '''numerical estimate''' of the required storage space based on '''projects''' that will use the allocation and their storage '''requirements''' .

'''Example 1''': "We will be processing a 1T dataset by performing 100 experimental runs.
Each experiment will be processed to produce a 6GB output, giving 600GB of the total output data.
We will also need 400GB additional space for post-processing and data management.
Thus, we would like to request '''2TB''' of shared space in total."

'''Example 2''': "3 members of our research group need additional shared space on ARC for their independent projects.
Project 1 starts with 100GB of initial data and is expected to generate 800GB of the output results.
Project 2 is going to use simulations and does not use any input data but is expected to generate 2TB of the simulated data for further processing.
The processing will require 200GB of additional space.
Project 3 will be working on a 1TB dataset and is expected to generate about 1TB of the output data.
These projects, therefore, will require 5.1TB of storage.
For convenience of data manipulation and management we would also like to have additional 400GB of extra storage space.
Therefore, we would like to request '''5.5TB''' of shared storage space in total."

Work and Bulk storage can be considerably larger than the home directory allocations. However, there are limits on what RCS can provide as ARC storage provides high-speed access and is expensive to purchase. Typically, '''any request over 10TB''' will require some discussion. Work and Bulk allocations differ in a few ways that influence how they are used. Work storage is faster to access as part of computational jobs on ARC although the impact is small for jobs that don't involve enormous numbers of reads. Bulk storage is designed to be a target for instrument data (which is typically processed in a way that reads data a small number of times per job) and is capable of mounting instruments elsewhere on campus using SMB. A number of questions come up frequently about Work and Bulk storage and these are addressed in an [[Group Storage Allocation FAQ | FAQ]].

===== How to a add group member to the access list (<code>/work</code> and <code>/bulk</code>)?=====

Any group member who wants to use the shared storage, should send an email to the support@hpc.ucalgary.ca to be added to the access group and CC the PI/ data owner. '''This will confirm that the PI approves the group member's request access to the shared storage.''' Please note that the access permissions inside the directory are expected to be managed by the data owners.

= Archive Storage =

[[Category:Administration]]
{{Navbox Administration}}

ARC Cluster Guide

2023-09-14T23:13:01Z

Darcy: /* ARC Cluster Storage */

{{ARC Cluster Status}}

{{Message Box
|title=[[Support|Need Help or have other ARC Related Questions?]]
|message=For all general RCS related issues, questions, or comments, please contact us at support@hpc.ucalgary.ca.
|icon=Support Icon.png}}

This guide gives an overview of the Advanced Research Computing (ARC) cluster at the University of Calgary and is intended to be read by new account holders getting started on ARC. This guide covers topics such as the hardware and performance characteristics, available software, usage policies and how to log in and run jobs. ARC can be used with data that a Researcher has classified as Lv1 and Lv2 as described in the UCalgary [https://www.ucalgary.ca/legal-services/sites/default/files/teams/1/Standards-Legal-Information-Security-Classification-Standard.pdf Information Security Classification Standard]

== Introduction ==
The ARC is a high performance compute (HPC) cluster that is available for research projects based at the University of Calgary. This compute cluster is comprised of hundreds of severs interconnected with a high bandwidth interconnect. Special resources within the cluster include nodes with large memory installed and GPUs are also available. You may learn more about ARC's hardware in the [[ARC Cluster Guide#Hardware|hardware section below]]. ARC can be accessed through a [[Linux Introduction|command line interface]] or via a web interface called Open OnDemand.

This cluster can be used for running large numbers (hundreds) of concurrent serial (one core) jobs, OpenMP or other thread-based jobs, shared-memory parallel code using up to 40 or 80 threads per job (depending on the partition), distributed-memory (MPI-based) parallel code using up to hundreds of cores, or jobs that take advantage of Graphics Processing Units (GPUs).

Historically, ARC is primarily comprised of older, disparate Linux-based clusters that were formerly offered to researchers from across Canada such as Breezy, Lattice, and Parallel. In addition, a large-memory compute node (Bigbyte) was salvaged from the now-retired local Storm cluster. In January 2019, a major addition to ARC with modern hardware was purchased. In 2020, compute clusters from CHGI have been migrated into ARC.

=== How to Get Started ===
If you have a project you think would be appropriate for ARC, please email support@hpc.ucalgary.ca and mention the intended research and software you plan to use. You must have a University of Calgary IT account in order to use ARC.
* For users that do not have a University of IT account or email address, please register for one at https://itregport.ucalgary.ca/.
* For users external to the University, such as for users collaborating on a research project at the University of Calgary, please contact us and mention the project leader you are collaborating with.

Once your access to ARC has been granted, you will be able to immediately make use of the cluster using your University of Calgary IT account by following the [[ARC_Cluster_Guide#Using_ARC|usage guide outlined below]].

== Using ARC ==

{{Message Box
|icon=Security Icon.png
|title=Cybersecurity awareness at the U of C
|message=Please note that there are typically about 950 phishing attempts targeting University of Calgary accounts each month. This is just a reminder to be careful about computer security issues, both at home and at the University. Please visit https://it.ucalgary.ca/it-security for more information, tips on secure computing, and how to report suspected security problems.}}

=== Logging in ===
To log in to ARC, connect using SSH to <code>arc.ucalgary.ca</code> on port <code>22</code>. Connections to ARC are accepted only from the University of Calgary network (on campus) or through the University of Calgary General VPN (off campus).

See [[Connecting to RCS HPC Systems]] for more information.
=== How to interact with ARC ===

ARC cluster is a collection of several compute nodes connected by a high-speed network. On ARC, computations get submitted as jobs. Once submitted, the jobs are then assigned to compute nodes by the job scheduler as resources become available.

[[File:Cluster.png]]

You can access ARC with your UCalgary IT user credentials. Once connected, you will get placed in the ARC login node, for basic tasks such as job submission, monitor job status, manage files, edit text, etc. It is a shared resource where multiple users get connected at the same time. Thus, any intensive tasks is not allowed on the login node as it may block other potential users to connect/submit their computations.
[tannistha.nandi@arc ~]$
The job scheduling system on ARC is called SLURM. On ARC, there are two SLURM commands that can allocate resources to a job under appropriate conditions: ‘salloc’ and ‘sbatch’. They both accept the same set of command line options with respect to resource allocation.

'''‘salloc’''' is to launch an interactive session, typically for tasks under 5 hours.
Once an interactive job session is created, you can do things like explore research datasets, start R or python sessions to test your code, compile software applications etc.

a. Example 1: The following command requests for 1 cpu on 1 node for 1 task along with 1 GB of RAM for an hour.
[tannistha.nandi@arc ~]$ salloc --mem=1G -c 1 -N 1 -n 1 -t 01:00:00
salloc: Granted job allocation 6758015
salloc: Waiting for resource configuration
salloc: Nodes fc4 are ready for job
[tannistha.nandi@fc4 ~]$

b. Example 2: The following command requests for 1 GPU to be used from 1 node belonging to the gpu-v100 partition along with 1 GB of RAM for 1 hour. Generic resource scheduling (--gres) is used to request for GPU resources.
[tannistha.nandi@arc ~]$ salloc --mem=1G -t 01:00:00 -p gpu-v100 --gres=gpu:1
salloc: Granted job allocation 6760460
salloc: Waiting for resource configuration
salloc: Nodes fg3 are ready for job
[tannistha.nandi@fg3 ~]$

Once you finish the work, type 'exit' at the command prompt to end the interactive session,
[tannistha.nandi@fg3 ~]$ exit
[tannistha.nandi@fg3 ~]$ salloc: Relinquishing job allocation 6760460
It is to ensure that the allocated resources are released from your job and now available to other users.

'''‘sbatch’''' is to submit computations as jobs to run on the cluster. You can submit a job-script.slurm via 'sbatch' for execution.
[tannistha.nandi@arc ~]$ sbatch job-script.slurm
When resources become available, they get allocated to this task. Batch jobs are suited for tasks that run for long periods of time without any user supervision. When the job-script terminates, the allocation is released.
Please review the section on how to prepare job scripts for more information.

=== Prepare job scripts ===
Job scripts are text files saved with an extension '.slurm', for example, 'job-script.slurm'.
A job script looks something like this:
''#!/bin/bash''
####### Reserve computing resources #############
#SBATCH --nodes=1
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=1
#SBATCH --time=01:00:00
#SBATCH --mem=1G
#SBATCH --partition=cpu2019 
####### Set environment variables ###############
module load python/anaconda3-2018.12 
####### Run your script #########################
python myscript.py

The first line contains the text "#!/bin/bash" to interpret it as a bash script.

It is followed by lines that start with a '#SBATCH' to communicate with 'SLURM'. You may add as many #SBATCH directives as needed to reserve computing resources for your task. The above example requests for one cpu on a single node for 1 task along with 1GB RAM for an hour on cpu2019 partition.

Next, you have to set up environment variables either by loading the modules centrally installed on ARC or export path to the software in your home directory. The above example loads an available python module.

Finally, include the Linux command to execute the local script.

Note that failing to specify part of a resource allocation request (most notably '''time''' and '''memory''') will result in bad resource requests as the defaults are not appropriate to most cases. Please refer to the section 'Running non-interactive jobs' for more examples.

== Hardware ==
Since the ARC cluster is a conglomeration of many different compute clusters, the hardware within ARC can vary widely in terms of performance and capabilities. To mitigate any compatibility issues with different hardware, we combine similar hardware into their own Slurm partition to ensure your workload runs as consistently as possible within one partition. Please carefully review the hardware specs for each of the partitions below to avoid any surprises.

=== Partition Hardware Specs ===
When submitting jobs to ARC, you may specify a partition that your job will run on. Please choose a partition that is most appropriate for your work.

* See also [[How to find available partitions on ARC]].

A few things to keep in mind when choosing a partition:
* Specific workloads requiring special Intel Instruction Set Extensions may only work on newer Intel CPUs.
* If working with multi-node parallel processing, ensure your software and libraries support the partition's interconnect networking.
* While older partitions may be slower, they may be less busy and have little to no wait times.

If you are unsure which partition to use or need assistance on selecting an appropriate partition, please see [[#Selecting_a_Partition|the Selecting a Partition Section]] below.

{| class="wikitable"
! Partition
! Description
! Nodes
! CPU Cores, Model, and Year
! Memory
! GPU
! Network
|-
| -
| ARC Login Node
| 1
| 16 cores, 2x Intel(R) Xeon(R) CPU E5620 @ 2.40GHz (Westmere, 2010)
| 48 GB
| N/A
| 40 Gbit/s InfiniBand
|-
| gpu-v100
| GPU Parition
| 13
| 80 cores, 2x Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz (Skylake, 2019)
| 754 GB
| 2x Tesla V100-PCIE-16GB
| 100 Gbit/s Omni-Path
|-
|gpu-a100
|GPU Partition
|5
|40 cores, 1x Intel(R) Xeon(R) Silver 4316 CPU @ 2.30GHz (Ice Lake, 2021)
|512 GB
|2x GA100 A100 PCIe 80GB
|100 Gbit/s Mellanox Infiniband
|-
|cpu2022
|General Purpose Compute
|52
|52 cores, 2x Intel(R) Xeon(R) Gold 5320 CPU @ 2.20GHz (Ice Lake)
|256 GB
|N/A
|40 Gbit/s InfiniBand
|-
| cpu2021
| General Purpose Compute
| 48
| 48 cores, 2x Intel(R) Xeon(R) Gold 6240R CPU @ 2.40GHz (Cascade Lake, 2021)
| 185 GB
| N/A
| 100 Gbit/s Mellanox Infiniband
|-
| cpu2019
| General Purpose Compute
| 14
| 40 cores, 2x Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz (Skylake, 2019)
| 190 GB
| N/A
| 100 Gbit/s Omni-Path
|-
| apophis
| General Purpose Compute
| 21
| 40 cores, 2x Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz (Skylake, 2019)
| 190 GB
| N/A
| 100 Gbit/s Omni-Path
|-
| razi
| General Purpose Compute
| 41
| 40 cores, 2x Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz (Skylake, 2019)
| 190 GB
| N/A
| 100 Gbit/s Omni-Path
|-
| bigmem
| Big Memory Nodes
| 2
| 80 cores, 4x Intel(R) Xeon(R) Gold 6148 CPU @ 2.40GHz (Skylake, 2019)
| 3022 GB
| N/A
| 100 Gbit/s Omni-Path
|-
| pawson
| General Purpose Compute
| 13
| 40 cores, 2x Intel(R) Xeon(R) Gold 6248 CPU @ 2.50GHz (Skylake, 2019)
| 190 GB
| N/A
| 100 Gbit/s Omni-Path
|-
|cpu2017
|General Purpose Compute
|14
|56 cores, 2x Intel(R) Xeon(R) CPU E5-2690 v4 @ 2.60GHz (Sandy Bridge, 2012)
|256 GB
|N/A
|40 Gbit/s InfiniBand
|-
| theia
| Former Theia cluster
| 20
| 56 cores, 2x Intel(R) Xeon(R) CPU E5-2690 v4 @ 2.60GHz (Sandy Bridge, 2012)
| 188 GB
| N/A
| 40 Gbit/s InfiniBand
|-
| cpu2013
| Former hyperion cluster
| 12
| 32 cores, 2x Intel(R) Xeon(R) CPU E5-2670 0 @ 2.60GHz (Sandy Bridge, 2012)
| 126 GB
| N/A
| 40 Gbit/s InfiniBand
|-
| lattice
| Former Lattice cluster
| 307
| 8 cores, 2x Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (Nehalem, 2009)
| 12 GB
| N/A
| 40 Gbit/s InfiniBand
|-
| single
| Former Lattice cluster
| 168
| 8 cores, 2x Intel(R) Xeon(R) CPU L5520 @ 2.27GHz (Nehalem, 2009)
| 12 GB
| N/A
| 40 Gbit/s InfiniBand
|-
| parallel
| Former Parallel Cluster
| 576
| 12 cores, 2x Intel(R) Xeon(R) CPU E5649 @ 2.53GHz (Westmere, 2011)
| 24 GB
| N/A
| 40 Gbit/s InfiniBand
|}

===ARC Cluster Storage===
Usage of ARC cluster storage is outlined by our [[ARC Storage Terms of Use]] page.

{{Warning Box
| title=Data Storage
| message=ARC storage is not suitable for long-term or archival storage. It is not backed-up and does not have sufficient redundancy to be used as a primary storage system. It is not guaranteed to be available for the time periods that are typical of archiving.

Please ensure that the only data you keep on ARC is used for active computations.

For information on available campus storage options, please see [[Storage Options]].
}}

{{Message Box
| title=No Backup Policy!
| message=You are responsible for your own backups. Many researchers will have accounts with Compute Canada and may choose to back up their data there (the Project file system accessible through the Cedar cluster would often be used).

Please contact us at support@hpc.ucalgary.ca if you want more information about this option.

You can also back up data to your UofC OneDrive for business allocation see: https://rcs.ucalgary.ca/How_to_transfer_data#rclone:_rsync_for_cloud_storage This allocation starts at 5TB. Contact the support center for questions regarding OneDrive for Business.
}}

The ARC cluster has around 2 petabyte of shared disk storage available across the entire cluster as well as temporary storage local to each of the compute nodes. Please refer to the individual sections below on the capacity limitations and usage policies.

Use the <code>arc.quota</code> command on ARC to determine the available space on your various volumes and home directory.

{| class="wikitable"
!Partition
!Description
!Capacity
|-
|<code>/home</code>
|User home directories
|500 GB (per user)
|-
|<code>/work</code>
|Research project storage
|Up to 100's of TB
|-
|<code>/scratch</code>
|Scratch space for temporary files
|Up to 15 TB
|-
|<code>/tmp</code>
|Temporary space local to the compute cluster
|Dependent on available storage on nodes. Verify with <code>df -h</code>.
|-
|<code>/dev/shm</code>
|Small temporary in-memory disk space local to the compute cluster
|Dependent on memory size set in your Slurm job.
|}
====<code>/home</code>: Home file system====
Each user has a directory under /home and is the default working directory when logging in to ARC. Each home directory has a per-user quota of 500 GB. This limit is fixed and cannot be increased. Researchers requiring additional storage exceeding what is available on their home directory may use <code>/work</code> and <code>/scratch</code>.

Note on file sharing: Due to security concerns, permissions set using <code>chmod</code> on your home directory to allow other users to read/write to your home directory be automatically reverted by an automated system process unless an explicit exception is made. If you need to share files with other researchers on the ARC cluster, please write to support@hpc.ucalgary.ca to ask for such an exception.

====<code>/scratch</code>: Scratch file system for large job-oriented storage====
Associated with each job, under the <code>/scratch</code> directory, a subdirectory is created that can be referenced in job scripts as <code>/scratch/${SLURM_JOB_ID}</code>. You can use that directory for temporary files needed during the course of a job. Up to 15 TB of storage may be used, per user (total for all your jobs) in the <code>/scratch</code> file system.

Data in <code>/scratch</code> associated with a given job will be deleted automatically, without exception, five days after the job finishes.

====<code>/work</code>: Work file system for larger projects====
If you need more space than provided in <code>/home</code> and the <code>/scratch</code> job-oriented space is not appropriate for you case, please write to support@hpc.ucalgary.ca with an explanation, including an indication of how much storage you expect to need and for how long. If approved, you will then be assigned a directory under <code>/work</code> with an appropriately large quota.

====<code>/tmp</code>,<code>/var/tmp</code>: Temporary files====
You may use <code>/tmp</code> or <code>/var/tmp</code> for storing temporary files generated by your job. The <code>/tmp</code> is stored on a disk local to the compute node and is not shared across the cluster. The files stored here will be removed immediately after your job terminates.

==== <code>/dev/shm</code>, <code>/run/user/$uid</code>: In-memory temporary files ====
<code>/dev/shm</code> and <code>/run/user/$UID</code> is writable location for temporary files backed by virtual memory. This can be used if faster I/O is required. This is ideal for workloads that require many small read/writes to share data between processes or as a fast cache. The amount of data you can write here is dependent on the amount of free memory available to your job. The files stored at these locations will be removed immediately after your job terminates.

== Software ==
All ARC nodes run the latest version of Rocky Linux 8 with the same set of base software packages. To maintain the stability and consistency of all nodes, any additional dependencies that your software requires must be installed under your account. For your convenience, we have packaged commonly used software packages and dependencies as modules available under <code>/global/software</code>. If your software package is not available as a module, you may also try Anaconda which allows users to manage and install custom packages in an isolated environment.

For a list of available packages that have been made available, please see [[ARC Software pages]].

Please contact us at support@hpc.ucalgary.ca if you need additional software installed.

==== Modules ====
The setup of the environment for using some of the installed software is through the <code>module</code> command. An overview of [https://www.westgrid.ca//support/modules modules on WestGrid (external link)] is largely applicable to ARC.

Software packages bundled as a module will be available under <code>/global/software</code> and can be listed with the <code>module avail</code> command.
<syntaxhighlight lang="bash">
$ module avail
</syntaxhighlight>

To enable Python, load the Python module by running:
<syntaxhighlight lang="bash">
$ module load python/anaconda-3.6-5.1.0
</syntaxhighlight>

To unload the Python module, run:
<syntaxhighlight lang="bash">
$ module remove python/anaconda-3.6-5.1.0
</syntaxhighlight>

To see currently loaded modules, run:
<syntaxhighlight lang="bash">
$ module list
</syntaxhighlight>

By default, no modules are loaded on ARC. If you wish to use a specific module, such as the Intel compilers or the Open MPI parallel programming packages, you must load the appropriate module.

== Job submission ==

=== Interactive Jobs ===
The ARC login node may be used for such tasks as editing files, compiling programs and running short tests while developing programs. We suggest CPU intensive workloads on the login node be restricted to under 15 minutes as per [[General Cluster Guidelines and Policies|our cluster guidelines]]. For interactive workloads exceeding 15 minutes, use the '''[[Running_jobs#Interactive_jobs|salloc command]]''' to allocate an interactive session on a compute node.

The default salloc allocation is 1 CPU and 1 GB of memory. Adjust this by specifying <code>-n CPU#</code> and <code>--mem Megabytes</code>. You may request up to 5 hours of CPU time for interactive jobs.
salloc --time=5:00:00 --partition=cpu2019

Always use salloc or srun to start an interactive job. Do not SSH directly to a compute node as SSH sessions will be refused without an active job running.



=== Running non-interactive jobs (batch processing) ===
Production runs and longer test runs should be submitted as (non-interactive) batch jobs, in which commands to be executed are listed in a script (text file). Batch jobs scripts are submitted using the <code>sbatch</code> command, part of the Slurm job management and scheduling software. #SBATCH directive lines at the beginning of the script are used to specify the resources needed for the job (cores, memory, run time limit and any specialized hardware needed).

Most of the information on the [https://docs.computecanada.ca/wiki/Running_jobs Running Jobs (external link)] page on the Compute Canada web site is also relevant for submitting and managing batch jobs and reserving processors for interactive work on ARC. One major difference between running jobs on the ARC and Compute Canada clusters is in selecting the type of hardware that should be used for a job. On ARC, you choose the hardware to use primarily by specifying a partition, as described below.

=== Selecting a Partition ===
There are some aspects to consider when selecting a partition including:
* Resource requirements in terms of memory and CPU cores
* Hardware specific requirements, such as GPU or CPU Instruction Set Extensions
* Partition resource limits and potential wait time
* Software support parallel processing using Message Passing Interface (MPI), OpenMP, etc.
** Eg. MPI for parallel processing can distribute memory across multiple nodes, per-node memory requirements could be lower. Whereas, OpenMP or single process code that is restricted to one node would require a higher memory node.
** Note: MPI code running on hardware with Omni-Path networking should be compiled with Omni-Path networking support. This is provided by loading the <code>openmpi/2.1.3-opa</code> or <code>openmpi/3.1.2-opa</code> modules prior to compiling.

Since resources that are requested are reserved for your job, please request only as much CPU and memory as your job requires to avoid reducing the cluster efficiency. If you are unsure which partition to use or the specific resource requests that are appropriate for your jobs, please contact us at [mailto:support@hpc.ucalgary.ca support@hpc.ucalgary.ca] and we would be happy to work with you.

{| class="wikitable" style="width: 100%;"
!Partition
!Description
!Cores/node
!Memory Request Limit
!Time Limit
!GPU
!Networking
|-
|cpu2021
|General Purpose Compute
|48
|185,000 MB
|7 days ‡
|
|100 Gbit/s Omni-Path
|-
|cpu2019
|General Purpose Compute
|40
|185,000 MB
|7 days ‡
|
|100 Gbit/s Omni-Path
|-
|bigmem
|Big Memory Compute
|80
|3,000,000 MB
|24 hours ‡
|
|100 Gbit/s Omni-Path
|-
|gpu-v100
|GPU Compute
|80
|753,000 MB
|24 hours ‡
|2
|100 Gbit/s Omni-Path
|-
|apophis&dagger;
|Private Research Partition
|40
|185,000 MB
|7 days ‡
|
|100 Gbit/s Omni-Path
|-
|razi&dagger;
|Private Research Partition
|40
|185,000 MB
|7 days ‡
|
|100 Gbit/s Omni-Path
|-
|pawson&dagger;
|Private Research Partition
|40
|185,000 MB
|7 days ‡
|
|100 Gbit/s Omni-Path
|-
|sherlock&dagger;
|Private Research Partition
|7
|185,000 MB
|7 days ‡
|
|100 Gbit/s Omni-Path
|-
|theia&dagger;
|Private Research Partition
|28
|188,000 MB
|7 days ‡
|
|40 Gbit/s InfiniBand
|-
|synergy&dagger;
|Private Research Partition
|14
|245,000 MB
|7 days ‡
|
|40 Gbit/s InfiniBand
|-
|cpu2013
|Legacy General Purpose Compute
|16
|120000
|7 days ‡
|
|40 Gbit/s InfiniBand
|-
|lattice
|Legacy General Purpose Compute
|8
|12000
|7 days ‡
|
|40 Gbit/s InfiniBand
|-
|parallel
|Legacy General Purpose Compute
|12
|23000
|7 days ‡
|
|40 Gbit/s InfiniBand
|-
|single
|Legacy Single-Node Job Compute
|8
|12000
|7 days ‡
|
|40 Gbit/s InfiniBand
|-
|cpu2021-bf24
|Back-fill Compute (2021-era hardware, 24h)
|48
|185,000 MB
|24 hours ‡
|
|100 Gbit/s Omni-Path
|-
|cpu2019-bf05
|Back-fill Compute (2019-era hardware, 5h)
|40
|185,000 MB
|5 hours ‡
|
|100 Gbit/s Omni-Path
|-
|cpu2017-bf05
|Back-fill Compute (2017-era hardware, 5h)
|14
|245,000 MB
|5 hours ‡
|
|40 Gbit/s InfiniBand
|-
|+ style="caption-side: bottom; text-align: left; font-weight: normal;" | &dagger; These partitions contain hardware contributed to ARC by particular researchers and should only be used by members of their research groups. However, they have generously allowed their compute nodes to be shared with others outside their research groups for short jobs. A special 'back-fill' or -bf partition is available for use by all ARC users for jobs shorter than 5 hours. ‡ As time limits may be changed by administrators to adjust to maintenance schedules or system load, the values given in the tables are not definitive. See the Time limits section below for commands you can use on ARC itself to determine current limits.
|}

==== Backfill partitions ====
Backfill partitions can be used by all users on ARC for short-term jobs. The hardware backing these partitions are generously contributed by researchers. We recommend including the backfill partitions for short term jobs as it may help reduce your job's wait time and increase the overall cluster throughput.

Previously, each contributing research group had their own backfill partition. Since June 2021, we have merged:

* apophis-bf, pawson-bf, and razi-bf into cpu2019-bf05
* theia-bf and synergy-bf into cpu2017-bf05

The naming scheme of the backfill partitions is the CPU generation year, followed by -bf and the time limit in hours. For example, cpu2017-bf05 would represent a backfill partition containing processors from 2017 with a time limit of 5 hours.

==== Hardware resource and job policy limits ====
In addition to the hardware limitations, please be aware that there may also be policy limits imposed on your account for each partition. These limits restrict the number of cores, nodes, or GPUs that can be used at any given time. Since the limits are applied on a partition-by-partition basis, using resources in one partition should not affect the available resources you can use in another partition.

These limits can be listed by running:
<syntaxhighlight lang="bash">
$ sacctmgr show qos format=Name,MaxWall,MaxTRESPU%20,MaxSubmitJobs
Name MaxWall MaxTRESPU MaxSubmit
---------- ----------- -------------------- ---------
normal 7-00:00:00 2000
breezy 3-00:00:00 cpu=384 2000
gpu 7-00:00:00 13000
cpu2019 7-00:00:00 cpu=240 2000
gpu-v100 1-00:00:00 cpu=80,gres/gpu=4 2000
single 7-00:00:00 cpu=408,node=75 2000
razi 7-00:00:00 2000
</syntaxhighlight>

==== Specifying a partition in a job ====
One you have decided which partitions best suits your computation, you can select one or more partition on a job-by-job basis by including the <code>partition</code> keyword for an <code>SBATCH</code> directive in your batch job. Multiple partitions should be comma separated. If you omit the partition specification, the system will try to assign your job to appropriate hardware based on other aspects of your request.

In some cases, you really should specify the partition explicitly. For example, if you are running single-node jobs with thread-based parallel processing requesting 8 cores you could use:
<syntaxhighlight lang="bash">
#SBATCH --mem=0 ❶
#SBATCH --nodes=1 ❷
#SBATCH --ntasks=1 ❸
#SBATCH --cpus-per-task=8 ❹
#SBATCH --partition=single,lattice ❺
</syntaxhighlight>

A few things to mention in this example:
# <code>--mem=0</code> allocates all available memory on the compute node for the job. This effectively allocates the entire node for your job.
# <code>--nodes=1</code> allocates 1 node for the job
# <code>--ntasks=1</code> your job has a single task
# <code>--cpus-per-task=8</code> asks for 8 CPUs per task. This job in total will request 8 * 1, or 8 CPUs.
# <code>--partition=single,lattice</code> specifies that this job can run on either single or lattice.
Suppose that your job requires at most 8 CPU cores and 10 GB of memory. The above Slurm request would be valid and optimal since your job fits neatly in a single node on the single and parallel partition. However, if you failed to specify the partition, Slurm may try to schedule your job to a partition with larger nodes, such as cpu2019 where each node has 40 cores and 190 GB of memory. If your job is scheduled on such a node, your job will be effectively wasting 32 cores and 180 GB of memory because <code>--mem=0</code> not only requests for 190 GB on this node, but also prevents other jobs from being scheduled on the same node.

If you don't specify a partition, please give greater thought to the memory specification to make sure that the scheduler will not assign your job more resources than are needed.

Parameters such as '''--ntasks-per-cpu''', '''--cpus-per-task''', '''--mem''' and '''--mem-per-cpu>''' have to be adjusted according to the capabilities of the hardware also. The product of --ntasks-per-cpu and --cpus-per-task should be less than or equal to the number given in the "Cores/node" column. The '''--mem>''' parameter (or the product of '''--mem-per-cpu''' and '''--cpus-per-task''') should be less than the "Memory limit" shown. If using whole nodes, you can specify '''--mem=0''' to request the maximum amount of memory per node.

===== Examples =====
Here are some examples of specifying the various partitions.

As mentioned in the [[#Hardware|Hardware]] section above, the ARC cluster was expanded in January 2019. To select the 40-core general purpose nodes specify:

#SBATCH --partition=cpu2019

To run on the Tesla V100 GPU-enabled nodes, use the '''gpu-v100''' partition. You will also need to include an SBATCH directive in the form '''--gres=gpu:n''' to specify the number of GPUs, n, that you need. For example, if the software you are running can make use of both GPUs on a gpu-v100 partition compute node, use:

#SBATCH --partition=gpu-v100 --gres=gpu:2

For very large memory jobs (more than 185000 MB), specify the bigmem partition:

#SBATCH --partition=bigmem

If the more modern computers are too busy or you have a job well-suited to run on the compute nodes described in the legacy hardware section above, choose the cpu2013, Lattice or Parallel compute nodes by specifying the corresponding partition keyword:

#SBATCH --partition=cpu2013
#SBATCH --partition=lattice
#SBATCH --partition=parallel

There is an additional partition called '''single''' that provides nodes similar to the lattice partition, but, is intended for single-node jobs. Select the single partition with

#SBATCH --partition=single

=== Time limits ===
Use the <code>--time</code> directive to tell the job scheduler the maximum time that your job might run. For example:
#SBATCH --time=hh:mm:ss

You can use <code>scontrol show partitions</code> or <code>sinfo</code> to see the current maximum time that a job can run.
<syntaxhighlight lang="bash" highlight="6">
$ scontrol show partitions
PartitionName=single
AllowGroups=ALL AllowAccounts=ALL AllowQos=ALL
AllocNodes=ALL Default=NO QoS=single
DefaultTime=NONE DisableRootJobs=NO ExclusiveUser=NO GraceTime=0 Hidden=NO
MaxNodes=UNLIMITED MaxTime=7-00:00:00 MinNodes=1 LLN=NO MaxCPUsPerNode=UNLIMITED
Nodes=cn[001-168]
PriorityJobFactor=1 PriorityTier=1 RootOnly=NO ReqResv=NO OverSubscribe=NO
OverTimeLimit=NONE PreemptMode=OFF
State=UP TotalCPUs=1344 TotalNodes=168 SelectTypeParameters=NONE
DefMemPerNode=UNLIMITED MaxMemPerNode=UNLIMITED
</syntaxhighlight>

Alternatively, with <code>sinfo</code> under the <code>TIMELIMIT</code> column:
<syntaxhighlight lang="bash">
$ sinfo
PARTITION AVAIL TIMELIMIT NODES STATE NODELIST
single up 7-00:00:00 1 drain* cn097
single up 7-00:00:00 1 maint cn002
single up 7-00:00:00 4 drain* cn[001,061,133,154]
...
</syntaxhighlight>

== Support ==
{{Message Box
|title=[[Support|Need Help or have other ARC Related Questions?]]
|message=For all general RCS related issues, questions, or comments, please contact us at support@hpc.ucalgary.ca.
|icon=Support Icon.png}}

Please don't hesitate to [[Support|contact us]] directly by email if you need help using ARC or require guidance on migrating and running your workflows to ARC.

[[Category:ARC]]
[[Category:Guides]]

2022-06-01T23:08:42Z

Darcy: /* Best Practices */

==Introduction==
CloudStack is an Infrastructure as a Service provided to University of Calgary researchers to allow them to quickly deploy Virtual Machines to support their research.

Researchers have a great degree of freedom in how they use CloudStack. As such, they are solely responsible for maintaining any VMs they deploy.

CloudStack is a research environment. While the system is available 24/7, support is only available during University business hours.

We ask that researchers follow the best practices listed below to ensure that the system remains available for the campus community.

==Best Practices==

# Please stay abreast of security updates for your OS and apply them.
# Backup your data (RCS does not provide backups).
# Do not run Windows (Our infrastructure is not licensed to run Windows).
# The University's "Acceptable Use of Electronic Resources and Information Policy" applies to your work using a VM. Please look [https://www.ucalgary.ca/legal-services/university-policies-procedures/ here] and search for "Electronic".
# This environment is only rated to handle Level 1 and Level 2 data. Please see [https://www.ucalgary.ca/legal-services/university-legal-services/operating-standards-guidelines-forms/ here] and select "Information Security Classification Standard" for details.
# If you create user accounts on your VM, make sure to manage them appropriately (remove/disable accounts that are no longer required).
# If you have created accounts, you are responsible for the appropriate use of the VM by those accounts.
# All user accounts on the VM must have good passwords. See [https://it.ucalgary.ca/it-security/passwords-do-i-have-change-them/ here] for details on creating strong passwords.
# CloudStack is not meant as a High Performance Computing (HPC) number cruncher. If you have HPC needs, please see "[[How to get an account]]" for details on how to apply for an account on ARC. Not sure what you need? Contact us at [mailto:support@hpc.ucalgary.ca support@hpc.ucalgary.ca].
# If your VM faces the outside world, please consider using appropriate security tools. Contact [mailto:support@hpc.ucalgary.ca support@hpc.ucalgary.ca] for assistance.

==Important notes==
In the event of a security incident, IT Operations/Security will shut down affected VMs.

Non-critical patches/upgrades to CloudStack that are required will happen on Tuesday of each week. Running VMs should not not be affected.

Non-critical patches/upgrades that require a complete restart of CloudStack will occur after 1 day email notice.

Urgent security patches to CloudStack may happen with little to no notice.

Please check back with this document as changes may occur.