Introduction to job scheduling

Job scheduler

An HPC system might have thousands of nodes and thousands of users. How do we decide who gets what and when? How do we ensure that a task is run with the resources it needs? This job is handled by a special piece of software called the scheduler. On an HPC system, the scheduler manages which jobs run, where and when.

The following illustration compares these tasks of a job scheduler to a waiter in a restaurant. If you can relate to an instance where you had to wait for a while in a queue to get into a popular restaurant, then you may now understand why sometimes your job does not start instantly as on your laptop.

The scheduler used here is Slurm. Although Slurm is not used everywhere, running jobs is quite similar regardless of what software is being used. The exact syntax might change, but the concepts remain the same.

Running a batch job

A basic script

The most basic use of the scheduler is to run a command non-interactively. Any command (or series of commands) that you want to run on the cluster is called a job and the process of using a scheduler to run the job is called batch job submission.

In this case, the job we want to run is a shell script – essentially a text file containing a list of UNIX commands to be executed in a sequential manner. Our shell script will have three parts:

• On the very first line, we’ll add #!/usr/bin/env bash. The #! (pronounced “hash-bang” or “shebang”) tells the computer what program is meant to process the contents of this file. In this case, we are telling it that the commands that follow are written for the command-line shell (what we’ve been doing everything in so far).
• Anywhere below the first line, we’ll add an echo command with a friendly greeting. When run, the shell script will print whatever comes after echo in the terminal.
  • echo -n will print everything that follows, without ending the line by printing the new-line character.
• On the last line, we’ll invoke the hostname command, which will print the name of the machine the script is run on.

Let’s use nano to write this script.

You can then use Ctrl-O followed by Enter to save the file and Ctrl-X to exit the editor.

Submitting the job

This script ran on the login node, but we want to take advantage of the compute nodes: we need the scheduler to queue up example-job.sh to run on a compute node.

To submit this task to the scheduler, we use the sbatch command. This creates a job which will run the script when dispatched to a compute node which the queuing system has identified as being available to perform the work.

However, running the script in its current form may yield an error like the following: 

In this case, it’s telling us that we need to specify more details for submitting it. As it turns out, as a minimum for COSMA we need to specify the following:

• Our account: this is the system’s budgetary account to which we individually are assigned. You can find out the accounts you have access to by using the sacctmgr list user yourUsername command.
• Partition: HPC system resources are typically split by type (such as cpu, gpu, large memory, etc.), or some other classification, into partitions. The configuration of these varies from system to system, but essentially are different queues to which you may submit your job. You may only be authorised to use specific partitions/queues.
• Walltime: the maximum amount of real-world time your job will take to run, specified as days-hours:minutes:seconds (the days can be omitted).

Note that depending on the system, other minimal parameters may also be necessary such as specifying a desired quality of service, or the minimum number of required nodes. But we’ll leave these for now.

We can specify them on the command line when submitting our job, like the following. Our job is very short running, so let’s just give it a maximum walltime of 1 minute:

And that’s what we need to do to submit a job. Our work is done; now the scheduler takes over and tries to run the job for us.

Monitoring our job

While the job is waiting to run, it goes into a list of jobs called the queue. To check on our job’s status, we check the queue using the command squeue:

Let’s try these out now:

We can see all the details of our job, including the partition, user, and also the state of the job (in the ST column). In this case, we can see it is in the PD or PENDING state. Typically, jobs go through the following states:

• PD – pending: sometimes our jobs might need to wait in a queue first before they can be allocated to a node to run.
• R – running: job has an allocation and is currently running.
• CG – completing: job is in the process of completing.
• CD – completed: the job is completed.

You can get a full list of job status codes via the SLURM documentation.

Customising a job

In the job we just ran we didn’t specify any detailed requirements for what our job will need. In a real-world scenario, that’s probably not what we want. Chances are, we will need more cores, more memory, more time, among other special considerations.

So far, we’ve specified job requirements directly on the command line which is quick and convenient, but somewhat limited and less clear when specifying many more parameters. Plus, we may forget which parameters we used in previous runs, which may be critical to reproducing a previous result. The good news is that we can amend our submission script directly to include these parameters instead.

Comments in UNIX shell scripts (denoted by #) are typically ignored, but there are exceptions. For instance the special #! comment at the beginning of scripts specifies what program should be used to run it (you’ll typically see #!/usr/bin/env bash). Schedulers like Slurm also have a special comment used to denote special scheduler-specific options. Though these comments differ from scheduler to scheduler, Slurm’s special comment is #SBATCH. Anything following the #SBATCH comment is interpreted as an instruction to the scheduler.

Let’s illustrate this by example. First, we’ll add the account, partition, and time parameters to the script directly, then give the job itself a different name. By default, a job’s name is the name of the script, but the --job-name (or -J for short) option can be used to change the name of a job. Amend the example-job.sh script to look like the following (amending yourAccount accordingly):

Submit the job and monitor its status: 

Fantastic, we’ve successfully changed the name of our job!

Resource requests

What about more important changes, such as the number of cores and memory for our jobs? One thing that is absolutely critical when working on an HPC system is specifying the resources required to run a job, which allows the scheduler to find the right time and place to schedule our job. If you do not specify requirements (such as the amount of time you need), you will likely be stuck with your site’s default resources, which is probably not what you want.

The following are several key resource requests:

• --ntasks=<ntasks> or -n <ntasks>: How many CPU cores does your job need, in total?

• --mem=<megabytes>: How much memory on a node does your job need in megabytes? You can also specify gigabytes using by adding a little “g” afterwards (example: --mem=5g).

• --nodes=<nnodes> or -N <nnodes>: How many separate machines does your job need to run on? Note that if you set ntasks to a number greater than what one machine can offer, Slurm will set this value automatically.

Note that just requesting these resources does not make your job run faster, nor does it necessarily mean that you will consume all of these resources. It only means that these are made available to you. Your job may end up using less memory, or less time, or fewer nodes than you have requested, and it will still run.

It’s best if your requests accurately reflect your job’s requirements. We’ll talk more about how to make sure that you’re using resources effectively in a later topic of this module.

Resource requests are typically binding. If you exceed them, your job will be killed. Let’s use walltime as an example. We will request 1 minute of walltime, and attempt to run a job for two minutes.

Submit the job and wait for it to finish. Once it has finished, check the log file.

Our job was killed for exceeding the amount of resources it requested. Although this appears harsh, this is actually a feature. Strict adherence to resource requests allows the scheduler to find the best possible place for your jobs. Even more importantly, it ensures that another user cannot use more resources than they’ve been given. If another user messes up and accidentally attempts to use all of the cores or memory on a node, Slurm will either restrain their job to the requested resources or kill the job outright. Other jobs on the node will be unaffected. This means that one user cannot mess up the experience of others, so the only jobs affected by a mistake in scheduling will be their own.

Cancelling a job

Sometimes we’ll make a mistake and need to cancel a job. This can be done with the scancel command. Let’s submit a job and then cancel it using its job number (remember to change the walltime so that it runs long enough for you to cancel it before it is killed!).

Now cancel the job with its job number (printed in your terminal). A clean return of your command prompt indicates that the request to cancel the job was successful.

Other types of jobs

Up to this point, we’ve focused on running jobs in batch mode. Slurm also provides the ability to start an interactive session.

There are very frequently tasks that need to be done interactively. Creating an entire job script might be overkill, but the amount of resources required is too much for a login node to handle. A good example of this might be building a genome index for alignment with a tool like HISAT2. Fortunately, we can run these types of tasks as a one-off with srun.

srun runs a single command on the cluster and then exits. Let’s demonstrate this by running the hostname command with srun. (We can cancel an srun job with Ctrl-C). Note that we still need to specify the account, partition, and expected run time as we would with any job, but in the case of srun, we can only specify these on the command line:

Note that given the interactive nature of the job, your terminal will pause until the job is able to be run, so you may have to wait. 

Interactive jobs

Sometimes, you will need a lot of resource for interactive use. Perhaps it’s our first time running an analysis, or we are attempting to debug something that went wrong with a previous job. Fortunately, Slurm makes it easy to start an interactive job with srun:

You should be presented with a bash prompt. Note that the prompt will likely change to reflect your new location, in this case the compute node we are logged on. You can also verify this with hostname.

When you are done with the interactive job, type exit to quit your session.