Using the NCI NeCTAR node
What is the NeCTAR Cloud?
NeCTAR is a federation of eight cloud nodes: NCI, QCIF, Intersect, Pawsey, ERSA, TPAC, Monash and the University of Melbourne (aka Lead Node), established with funding from the Australian Government through one of its Super Science Initiatives.
The NCI NeCTAR node
- 3200 cores (HT)
- 200 compute nodes
- 2 * GPU nodes
- 16C/128GB ram/2x400G SSD
- Intel Xeon CPU E5-2670 @ 2.60GHz
- 13 * Ceph storage nodes
- 56 GbE (Mellanox) network
Images and Instances
The NeCTAR research cloud provides resources by way of virtual machines. They can be treated just like real-machines with operating system, network access and disk storage but built-for-purpose with arbitrary lifetime.
- Images are disk images which are templates for virtual machine file systems. Details for images provided by NeCTAR can be found in the Image Catalog.
- Instances are the individual virtual machines running on physical compute nodes. They originate from Images and can be a plain “off the shelf” Operating System or include software packages and config changes to suit a particular purpose (e.g. webserving). Any number of instances maybe started from the same image.
- When starting an instance a set of virtual resources known as a flavor must be selected. Flavors define how many virtual CPUs an instance has and the amount of RAM and size of its ephemeral disks.
Requesting resources on the NeCTAR Research Cloud
You can run instances of various sizes on the cloud, from one to 16 cores, and from one instance to hundreds.
Project Trials (Automatic, no application required)
When you log into the cloud for the first time, you are automatically granted a Project Trial Research Allocation of two cores for three months.
Within your default allocation you can run:
- a medium (two core) instance, or
- two small (single core) instances.
To submit a request for more resources use the Allocation Request form from the NeCTAR Dashboard.
As a rule of thumb, the more resources you ask for, the more detail is required about your research. Requesting a few cores won’t be scrutinised as much as requesting tens or hundreds of cores.
Support for the NeCTAR Research Cloud is handled by NeCTAR and available weekdays 9am AEST to 5pm AWST through their website.
- First read the known issues and FAQ for known problems.
- Search the site: forum posts and support tickets may have answers or show an issue is already reported.
- See if your issue is supported (“Issues Support Staff will respond to“)
Otherwise, email a Support Request following the instructions at http://support.rc.nectar.org.au/docs/getting-support. You may also get answers at the NeCTAR Community Forum.
Exercise: Getting started on the NeCTAR cloud
Access the NeCTAR cloud through: http://dashboard.rc.nectar.org.au/
NeCTAR uses the Australian Access Federation (AAF) for its authentication.
Australian National University as your organisation and login with your ANU Uni ID and password.
You should now be logged in to the NeCTAR Research Cloud Dashboard
Create a Keypair
- Generate an SSH keypair, eg on your Linux or Mac OSX computer
ssh-keygen -t rsa -f cloud.key
- This will create 2 files, a public key and a private key:
- For Windows users use the
- Go to Access and Security Tab on the Dashboard and then to the Keypairs page
- Click ‘Import Keypair’
- Type in a keypair name that you can remember it by and paste the contents of
cloud.key.pubor the PuTTY Key fingerprint
Boot an instance
- Go to the Dashboard Images tab
- Select a NeCTAR offical image, such as NeCTAR Centos 6.5 x86_64 (that is closest to what is on raijin)
- Click on launch
- Give the instance a name (of your choosing again)
- Select a flavour (the default of m1.small will do)
- Under Availability Zone you will see the list of nodes – select NCI.
- Under Access and Security select your Keypair created earlier.
- Here you can also select what security groups to use. Select ‘ssh’, ‘http’ and ‘icmp’ (these pre-defined groups are available in Project Trials only)
- Click ‘Launch’… the page will change to the Instances Tab
- After a while (usually a couple of minutes) you instance will become active and will have an IP address
- Now you should be able to SSH into your new instance
ssh -i cloud.key ec2-user@XX.XX.XX.XX
Note that other images may expect other usernames. Try using
root@XX.XX.XX.XX for other Operating Systems.
- Before you do anything make sure you run a
yum updateas there may be security updates released between image builds.
sudo yum update
Configure your web server
sudo yum install httpd
sudo service httpd start
sudo vi /etc/httpd/conf.d/welcome.conf– to comment out the redirect lines
sudo vi /var/www/html/index.html– to add a welcoming message e.g.
Welcome to my cloud!
- Now copy and paste your IP address into your browser
Terminating your instance
When you finish with your instance don’t forget to Terminate it. This is the equivalent of destroying the machine and cannot be reversed. If you want to return to the instance at a later time you can take a copy of it with “Create Snapshot”.