What resources are available on the cluster?

Overview

Teaching: 0 min
Exercises: 0 min

Questions

• What resources can I request on a cluster?

• How do I find out which resources are available on the cluster I’m using?

Objectives

• Use the command line to determine the resources available on the cluster.

• Use other tools and resources to determine the resources available on the cluster.

So far, you’ve learned

• What a cluster is
• Why you might want to use one
• How to log into one.

I’ve told you that there are different resources on a HPC than in your local workstation, and in this section I’ll show you how to find out what’s available to you.

What resources are available in general?

There are a variety of computational resources you may be aware of. These include:

• CPUs (Central Processing Units). These are designed for sequential processing.
• GPUs (Graphical Processing Units). These are traditionally designed for displaying graphics. Graphics are represented with matrices, meaning GPU architecture is excellent at performing matrix multiplications. Some HPC clusters will have access to “data center” GPUs, especially designed for computation rather than graphics. These will differ from what’s typically available to your local workstation.
• TPUs (Tensor Processing Units). These processors have been especially designed for working with tensors and neural networks, which you may be familiar with from TensorFlow or PyTorch. Notably, Google Colab provides access to TPUs, with an introductory tutorial notebook.
• Time. When using a HPC you will need to specify how long you want to use the shared resources for.
• RAM (Random Access Memory). You will need to consider how much memory is required to run your code.
• Software. Every computer will come with its own unique set of installed software - the HPC is included. We’ll talk more about specific software later.

Local resources

Take a moment to think about what resources are available on your local workstation. How many CPUs? Do you have a GPU? Any specialised software?

How do I know what’s available on MASSIVE?

There are a variety of methods to discover which resources are available to you on a given cluster. Most HPC sites will offer a documentation website - for MASSIVE, you can find our documentation on docs.massive.org.au. In particular, you can find out more about our hardware on the About M3 page, or about our GPUs specifically in our section on GPUs.

We also provide some of this documentation in the Strudel2 interface. Under the Account Info tab, you are able to see a list of desktop types that are available, and under the Desktops tab, you will be able to learn more about the specific resources assigned to each desktop type.

Another method to find out information about the resources on the cluster is using the command line. You will remember from “What is a cluster?” that we group nodes with the same resources into partitions. Each of these nodes will have a particular set of resources like CPUs and RAM associated with it. There are a few ways to learn more about these partitions and nodes on the command line that we’ll explore here. Most of these commands are associated with SLURM, the scheduler for our HPC.

What partitions are available?

• sinfo -s : sinfo is a SLURM command that gives you information about nodes and partitions. This particular command summarises the partitions, where the -s flag indicates summary. It will output a list of each partition, whether it’s available at the moment (and not down for maintenance, for example), the time limits or wall time for using that partition, how many nodes are allocated, idle, out, and total, and which nodes belong to the partition.
• scontrol show partitions : scontrol is a SLURM command for configuring SLURM, and this command gives you a more in depth view over the partitions available and how they have been configured. There may be information in this ouput which you don’t understand - this is okay, for our purposes it’s another way of seeing which partitions are on MASSIVE. You can also use scontrol show partition <partition-name> to investigate specific partitions.

  What nodes are available?

• sinfo --partition=<partition-name> -N --format=%10N%10c%10m%20f%10G : This sinfo command gives more details about the nodes on a partition. It will print out a list of node names, the number of CPUs, the memory on the node, and any other features such as GPUs.
• show_cluster : show_cluster is a command developed for MASSIVE which presents a formatted summary of the status of nodes on M3. For a similar output on other clusters, you can use sinfo -lN.

Investigating the Cluster

1. Try typing in the commands above and inspect their output.
2. What type of GPUs are there on the m3g partition? How did you find out?

Solution

1. There are a variety of formats with different information available to you for learning about the cluster!
2. There are V100 GPUs on the m3g partition. You could have learned this by running:
   • sinfo --partition=m3g -N --format=%10N%10c%10m%20f%10G
   • show_cluster
   • sinfo -lN
   • Checking docs.massive.org.au for more information

As you can see, there are many ways to learn about the cluster!

Key Points

• Different clusters will have different resources available on them.

• You can check resources on the cluster with the command line, or by using documentation.