Quickstart new

Niagara
Installed	Jan 2018
Operating System	CentOS 7.4
Number of Nodes	1500 nodes (60,000 cores)
Interconnect	Mellanox Dragonfly+
Ram/Node	188 GiB / 202 GB
Cores/Node	40 (80 hyperthreads)
Login/Devel Node	niagara.scinet.utoronto.ca
Vendor Compilers	icc (C) ifort (fortran) icpc (C++)
Queue Submission	Slurm

Specifications

The Niagara cluster is a large cluster of 1500 Lenovo SD350 servers each with 40 Intel "Skylake" cores at 2.4 GHz. The peak performance of the cluster is 3.02 PFlops delivered / 4.6 PFlops theoretical. It is the 53rd fastest supercomputer on the TOP500 list of June 2018.

Each node of the cluster has 188 GiB / 202 GB RAM per node (at least 4 GiB/core for user jobs). Being designed for large parallel workloads, it has a fast interconnect consisting of EDR InfiniBand in a Dragonfly+ topology with Adaptive Routing. The compute nodes are accessed through a queueing system that allows jobs with a minimum of 15 minutes and a maximum of 12 or 24 hours and favours large jobs.

• See the "Intro to Niagara" recording

More detailed hardware characteristics of the Niagara supercomputer can be found on this page.

Getting started on Niagara

Those of you new to SciNet and belonging to a group whose primary PI does not have an allocation, as granted in the annual Compute Canada RAC, must first follow the old route of requesting a SciNet Consortium Account on the CCDB site to gain access to Niagara.

Please read this document carefully. The FAQ is also a useful resource. If at any time you require assistance, or if something is unclear, please do not hesitate to contact us.

Logging in

Niagara runs CentOS 7, which is a type of Linux. You will need to be familiar with Linux systems to function on Niagara. If you are not it will be worth your time to review the our Introduction to Linux Shell class.

As with all SciNet and CC (Compute Canada) compute systems, access to Niagara is done via SSH (secure shell) only. Open a terminal window (e.g. Connecting with PuTTY on Windows or Connecting with MobaXTerm), then SSH into the Niagara login nodes with your CC credentials:

$ ssh -Y MYCCUSERNAME@niagara.scinet.utoronto.ca

or

$ ssh -Y MYCCUSERNAME@niagara.computecanada.ca

• The Niagara login nodes are where you develop, edit, compile, prepare and submit jobs.
• These login nodes are not part of the Niagara compute cluster, but have the same architecture, operating system, and software stack.
• The optional -Y is needed to open windows from the Niagara command-line onto your local X server.
• To run on Niagara's compute nodes, you must submit a batch job.

If you cannot log in, be sure first to check the System Status on this site's front page.

Your various directories

By virtue of your access to Niagara you are granted storage space on the system. There are several directories available to you, each indicated by an associated environment variable.

home and scratch

You have a home and scratch directory on the system, whose locations will be given in the form

$HOME=/home/g/groupname/myccusername
$SCRATCH=/scratch/g/groupname/myccusername

where groupname is the name of your PI's group, and myccusername is your CC username. For example:

 nia-login07:~$ pwd
 /home/s/scinet/rzon
 nia-login07:~$ cd $SCRATCH
 nia-login07:rzon$ pwd
 /scratch/s/scinet/rzon

NOTE: home is read-only on compute nodes.

project and archive

Users from groups with RAC storage allocation will also have a project and/or archive directory.

$PROJECT=/project/g/groupname/myccusername
$ARCHIVE=/archive/g/groupname/myccusername

NOTE: Currently archive space is available only via HPSS.

IMPORTANT: Future-proof your scripts

When writing your scripts, use the environment variables ($HOME, $SCRATCH, $PROJECT, $ARCHIVE) instead of the actual paths! The paths may change in the future.

Storage and quotas

You should familiarize yourself with the various file systems, what purpose they serve, and how to properly use them. This table summarizes the various file systems. See the Data Management page for more details.

location	quota		block size	expiration time	backed up	on login nodes	on compute nodes
$HOME	100 GB per user		1 MB		yes	yes	read-only
$SCRATCH	25 TB per user		16 MB	2 months	no	yes	yes
	50-500TB per group	depending on group size
$PROJECT	by group allocation		16 MB		yes	yes	yes
$ARCHIVE	by group allocation				dual-copy	no	no
$BBUFFER	10 TB per user		1 MB	very short	no	yes	yes

Moving data to Niagara

If you need to move data to Niagara for analysis, or when you need to move data off of Niagara, use the following guidelines:

• If your data is less than 10GB, move the data using the login nodes.
• If your data is greater than 10GB, move the data using the datamover nodes nia-datamover1.scinet.utoronto.ca and nia-datamover2.scinet.utoronto.ca .

Details of how to use the datamover nodes can be found on the Data Management page.

File Input/Output Tips

It is important to understand the file systems, so as to perform your file I/O (Input/Output) responsibly. Refer to the Data Management page for details about the file systems.

• Your files can be seen on all Niagara login and compute nodes.
• $HOME, $SCRATCH, and $PROJECT all use the parallel file system called GPFS.
• GPFS is a high-performance file system which provides rapid reads and writes to large data sets in parallel from many nodes.
• Accessing data sets which consist of many, small files leads to poor performance on GPFS.
• Avoid reading and writing lots of small amounts of data to disk. Many small files on the system waste space and are slower to access, read and write. If you must write many small files, use ramdisk.
• Write data out in a binary format. This is faster and takes less space.
• The Burst Buffer is another option for I/O heavy-jobs and for speeding up checkpoints.

Progressive approach to test and run jobs on niagara

1) Choose the proper software stack: NiaEnv(default) or CCEnv.

NiaEnv and CCEnv

On Niagara, there are really two software stacks:

1. A Niagara software stack tuned and compiled for this machine. This stack is available by default, but if not, can be reloaded with

   module load NiaEnv

2. The same software stack available on Compute Canada's General Purpose clusters Graham and Cedar, compiled (for now) for a previous generation of CPUs:

   module load CCEnv

   Or, if you want the same default modules loaded as on Cedar and Graham, then do

   module load CCEnv

   module load StdEnv

2) Determine whether you will use existing software on Niagara or compile your own executable on the login nodes.

Using existing software

Other than essentials, all installed software is made available using module commands. These modules set environment variables (PATH, etc.), allowing multiple, conflicting versions of a given package to be available. A detailed explanation of the module system can be found on the modules page.

Common module subcommands are:

• module load <module-name>: load the default version of a particular software.
• module load <module-name>/<module-version>: load a specific version of a particular software.
• module purge: unload all currently loaded modules.
• module spider (or module spider <module-name>): list available software packages.
• module avail: list loadable software packages.
• module list: list loaded modules.

Along with modifying common environment variables, such as PATH, and LD_LIBRARY_PATH, these modules also create a SCINET_MODULENAME_ROOT environment variable, which can be used to access commonly needed software directories, such as /include and /lib.

There are handy abbreviations for the module commands. ml is the same as module list, and ml <module-name> is the same as module load <module-name>.

Compiling on Niagara: Example

Suppose one want to compile an application from two c source files, appl.c and module.c, which use the Math Kernel Library. This is an example of how this would be done:

nia-login07:~$ module list
Currently Loaded Modules:
  1) NiaEnv/2018a (S)
  Where:
   S:  Module is Sticky, requires --force to unload or purge

nia-login07:~$ module load intel/2018.2

nia-login07:~$ ls
appl.c module.c

nia-login07:~$ icc -c -O3 -xHost -o appl.o appl.c
nia-login07:~$ icc -c -O3 -xHost -o module.o module.c
nia-login07:~$ icc  -o appl module.o appl.o -mkl

nia-login07:~$ ./appl

Note:

• The optimization flags -O3 -xHost allow the Intel compiler to use instructions specific to the architecture CPU that is present (instead of for more generic x86_64 CPUs).
• Linking with the Intel Math Kernel Library (MKL) is easy when using the intel compiler, it just requires the -mkl flags.
• If compiling with gcc, the optimization flags would be -O3 -march=native. For the way to link with the MKL, it is suggested to use the MKL link line advisor.

Tips for loading software

• We advise against loading modules in your .bashrc. This can lead to very confusing behaviour under certain circumstances. Our guidelines for .bashrc files can be found here.
• Instead, load modules by hand when needed, or by sourcing a separate script.
• Load run-specific modules inside your job submission script.
• Short names give default versions; e.g. intel → intel/2018.2. It is usually better to be explicit about the versions, for future reproducibility.
• Modules often require other modules to be loaded first. Solve these dependencies by using module spider.

3) Test you code, by making sure the executable runs on a login node at least over 1 CPU, then 2 CPUs and then 4 CPUs max, and for no more then 15 minutes. These are only preliminary checks, not production runs.

Testing

You really should test your code before you submit it to the cluster to know if your code is correct and what kind of resources you need.

• Small test jobs can be run on the login nodes. Rule of thumb: tests should run no more couple of minutes, taking at most about 1-2GB of memory, and use no more than a couple of cores.
• You can run the the ddt debugger on the login nodes after module load ddt.

4) from this point on, request an interactive session on the debug queue for *ONE* node, at most 1 hour, and ensure you can scale up to all 40 CPUs, without running out of memory or some other single node specific hiccup. Pay attention to all messages and notifications on the standard output, and fix all the bugs detected up to this stage, if any.

• Short tests that do not fit on a login node, or for which you need a dedicated node, request an interactive debug job with the debug command:

nia-login07:~$ debugjob N

where N is the number of nodes, If N=1, this gives an interactive session one 1 hour, when N=4 (the maximum), it gives you 30 minutes. Finally, if your debugjob process takes more than 1 hour, you can request an interactive job from the regular queue using the salloc command. Note, however, that this may take some time to run, since it will be part of the regular queue, and will be run when the scheduler decides.

nia-login07:~$ salloc --nodes N --time=M:00:00

where N is again the number of nodes, and M is the number of hours you wish the job to run. If you need to use graphics while testing your code through salloc, e.g. when using a debugger such as DDT or DDD, you have the following options, please visit the Testing with graphics page.

5) then request an interactive session on the debug queue for *TWO* nodes, at most 1 hour, start adjusting your scripts to run over multiple nodes, possibly try hyperthreading.

6) then focus on the submission script itself in batch mode, submit it to 1 node on the debug queue, for 15 minutes, then for 2 nodes for 30 minutes, login to the nodes to check what is going on, and again, pay close attention to any error messages or notifications on the logs.

Example submission script (MPI)

#!/bin/bash 
#SBATCH --nodes=2
#SBATCH --ntasks=80
#SBATCH --time=1:00:00
#SBATCH --job-name mpi_job
#SBATCH --output=mpi_output_%j.txt

cd $SLURM_SUBMIT_DIR

module load intel/2018.2
module load openmpi/3.1.0

mpirun ./mpi_example
# or "srun ./mpi_example"

Submit this script with the command:

   nia-login07:~$ sbatch mpi_job.sh

• First line indicates that this is a bash script.
• Lines starting with #SBATCH go to SLURM.
• sbatch reads these lines as a job request (which it gives the name mpi_job)
• In this case, SLURM looks for 2 nodes (each of which will have 40 cores) on which to run a total of 80 tasks, for 1 hour.
  (Instead of specifying --ntasks=80, you can also ask for --ntasks-per-node=40, which amounts to the same.)
• Note that the mpifun flag "--ppn" (processors per node) is ignored.
• Once it found such a node, it runs the script:
  • Change to the submission directory;
  • Loads modules;
  • Runs the mpi_example application (SLURM will inform mpirun or srun on how many processes to run).
• To use hyperthreading, just change --ntasks=80 to --ntasks=160, and add --bind-to none to the mpirun command (the latter is necessary for OpenMPI only, not when using IntelMPI).

Example submission script (OpenMP)

#!/bin/bash
#SBATCH --nodes=1
#SBATCH --cpus-per-task=40
#SBATCH --time=1:00:00
#SBATCH --job-name openmp_job
#SBATCH --output=openmp_output_%j.txt

cd $SLURM_SUBMIT_DIR

module load intel/2018.2

export OMP_NUM_THREADS=$SLURM_CPUS_PER_TASK

./openmp_example
# or "srun ./openmp_example".

Submit this script with the command:

   nia-login07:~$ sbatch openmp_job.sh

• First line indicates that this is a bash script.
• Lines starting with #SBATCH go to SLURM.
• sbatch reads these lines as a job request (which it gives the name openmp_job) .
• In this case, SLURM looks for one node with 40 cores to be run inside one task, for 1 hour.
• Once it found such a node, it runs the script:
  • Change to the submission directory;
  • Loads modules;
  • Sets an environment variable;
  • Runs the openmp_example application.
• To use hyperthreading, just change --cpus-per-task=40 to --cpus-per-task=80.

7) From this point on, and only then, submit your jobs to the normal batch queue, for *ONE* hour at first and for *TWO* nodes only. Everything going well start to introduce checkpoint strategies to your dataset and workflow, so you may pickup the slack in case of disruptions on the execution.

8) after you developed a reasonable checkpoint procedure, and only then, slowly scale up your submission in 2 dimensions, 4, 10 and 20 nodes, then 4, 8 and 12 hours. Be sure your type of job scales well over higher number of nodes first, before asking for more time.

Once you have compiled and tested your code or workflow on the Niagara login nodes, and confirmed that it behaves correctly, you are ready to submit jobs to the cluster. Your jobs will run on some of Niagara's 1500 compute nodes. When and where your job runs is determined by the scheduler.

Niagara uses SLURM as its job scheduler. More-advanced details of how to interact with the scheduler can be found on the Slurm page.

You submit jobs from a login node by passing a script to the sbatch command:

nia-login07:~$ sbatch jobscript.sh

This puts the job in the queue. It will run on the compute nodes in due course.

Jobs will run under the user's group's RRG allocation, or, if the group has none, under a RAS allocation (previously called `default' allocation).

Keep in mind:

• Scheduling is by node, so in multiples of 40 cores.
• For users with an allocation, the maximum walltime is 24 hours. For those without an allocation, the maximum walltime is 12 hours.
• Jobs must write their output to your scratch or project directory (home is read-only on compute nodes).
• Compute nodes have no internet access.
• Move your data to Niagara before you submit your job.

Scheduling by Node

On many systems that use SLURM, the scheduler will deduce from the specifications of the number of tasks and the number of cpus-per-node what resources should be allocated. On Niagara things are a bit different.

• All job resource requests on Niagara are scheduled as a multiple of nodes.
• The nodes that your jobs run on are exclusively yours, for as long as the job is running on them.
  • No other users are running anything on them.
  • You can SSH into them to see how things are going.
• Whatever your requests to the scheduler, it will always be translated into a multiple of nodes allocated to your job.
• Memory requests to the scheduler are of no use. Your job always gets N x 202GB of RAM, where N is the number of nodes and 202GB is the amount of memory on the node.
• If you run serial jobs you must still use all 40 cores on the node. Visit the serial jobs page for examples of how to do this.
• Since there are 40 cores per node, your job should use N x 40 cores. If you do not, we will contact you to help you optimize your workflow. Or you can contact us to get assistance.

Limits

There are limits to the size and duration of your jobs, the number of jobs you can run and the number of jobs you can have queued. It matters whether a you are part of a group with a Resources for Research Group allocation or not. It also matters in which 'partition' the job runs. 'Partitions' are SLURM-speak for use cases. You specify the partition with the -p parameter to sbatch or salloc, but if you do not specify one, your job will run in the compute partition, which is the most common case.

Usage	Partition	Limit on Running jobs	Limit on Submitted jobs (incl. running)	Min. size of jobs	Max. size of jobs	Min. walltime	Max. walltime
Compute jobs with an allocation	compute	50	1000	1 node (40 cores)	1000 nodes (40000 cores)	15 minutes	24 hours
Compute jobs without allocation ("default")	compute	50	200	1 node (40 cores)	20 nodes (800 cores)	15 minutes	12 hours
Testing or troubleshooting	debug	1	1	1 node (40 cores)	4 nodes (160 cores)	N/A	1 hour
Archiving or retrieving data in HPSS	archivelong	2 per user (max 5 total)	10 per user	N/A	N/A	15 minutes	72 hours
Inspecting archived data, small archival actions in HPSS	archiveshort	2 per user	10 per user	N/A	N/A	15 minutes	1 hour

Even if you respect these limits, jobs will still have to wait in the queue. The waiting time depends on many factors such as the allocation amount, how much allocation was used in the recent past, the number of nodes and the walltime, and how many other jobs are waiting in the queue.

Monitoring queued jobs

Once the job is incorporated into the queue, there are some command you can use to monitor its progress.

• squeue or sqc (a caching version of squeue) to show the job queue (squeue -u $USER for just your jobs);

• squeue -j JOBID to get information on a specific job

  (alternatively, scontrol show job JOBID, which is more verbose).

• squeue --start -j JOBID to get an estimate for when a job will run; these tend not to be very accurate predictions.

• scancel -i JOBID to cancel the job.

• jobperf JOBID to get an instantaneous view of the cpu and memory usage of the nodes of the job while it is running.

• sacct to get information on your recent jobs.

Further instructions for monitoring your jobs can be found on the Slurm page. The my.SciNet site is also a very useful tool for monitoring your current and past usage.

Visualization

Information about how to use visualization tools on Niagara is available on Visualization page.

Visualization

Information about how to use visualization tools on Niagara is available on Visualization page.

PLEASE, REFRAIN FROM SUBMITTING A JOB FOR THE FIRST TIME ASKING ALREADY FOR THE MAX NUMBER OF NODES AND THE MAX AMOUNT OF TIME, IF YOU ARE NOT REASONABLY SURE IT WILL WORK, OR YOU WILL BE WASTING A LOT OF RESOURCES.