The primary way to use the NCI cloud is via the web dashboard. To log in, follow these steps:
You will be taken to your default project overview page. If you are a member of multiple cloud projects you can select the project to view from the drop-down box on the top left of the screen.
If you are using Nirin for the first time, ensure at least one key pair exists under your account.
something.pub
, often id_rsa.pub
in your ~/.ssh
directory).Setting up a single key pair is a one-time operation. This is the minimum requirement to get started using this guide; more complex workflows or security models may require multiple key pairs.
Creating an instance in the dashboard is done using the Launch Instance wizard, which takes you through a number of dialogs collecting information about the instance to be created. Follow the steps below to create a basic instance.
Once the Status becomes "Active" the instance is up and running.
The instance created in the previous section has a network connection with a private IP address (10.0.0.18), and can connect to hosts on the Internet, but it is not accessible from the wider Internet. In order to access it externally you need to attach a floating IP to it - this is an externally accessible address that will forward traffic through to the instance on its private address. Follow the steps below to allocate a floating IP and attach it to the instance created previously.
ssh ubuntu@130.56.246.60
.When a floating IP address is no longer required, it can be released back to the pool by pressing "Release Floating IP" in the "Actions" column.
Security risk: releasing floating IP addresses
Releasing a floating IP address makes it available for other projects to use, which can in some situations present a security risk. For example, if an external system (such as an old NFS export of gdata filesystems) refers to a public IP address which is released back to the pool and later re-used by another project on the cloud, this could inadvertently give that project access to the external system. To mitigate this risk: before releasing a floating IP address, ensure that any external systems that rely on it are told that the address is no longer valid.
Instances in the cloud have a small (10GB) local disk that is intended to hold only the operating system install and its essential files. Additional storage is provided through the use of volume storage. Volumes are effectively an external drive that can be attached to and detached from the instance as required - the volume continues to exist even after the instance is deleted.
To create a volume and attach it to an instance follow these steps:
Log in to your instance and run the lsblk
command to list the block devices. You should see your volume presented on the device listed previously.
ubuntu@my-instance:~$ lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT vda 252:0 0 10G 0 disk ├─vda1 252:1 0 9.9G 0 part / ├─vda14 252:14 0 4M 0 part └─vda15 252:15 0 106M 0 part /boot/efi vdb 252:16 0 1G 0 disk
At this point the volume can have a filesystem created on it, and it can be mounted and used.
ubuntu@my-instance:~$ sudo mkfs.ext4 /dev/vdb mke2fs 1.44.1 (24-Mar-2018) Creating filesystem with 262144 4k blocks and 65536 inodes Filesystem UUID: 28813bd6-e50a-40ef-864b-9f4aaa541eac Superblock backups stored on blocks: 32768, 98304, 163840, 229376 Allocating group tables: done Writing inode tables: done Creating journal (8192 blocks): done Writing superblocks and filesystem accounting information: done ubuntu@my-instance:~$ sudo mkdir /data ubuntu@my-instance:~$ sudo mount /dev/vdb /data ubuntu@my-instance:~$ sudo touch /data/test ubuntu@my-instance:~$ ls /data/test /data/test
NCI offers a range of virtual machine flavours to suit a range of needs. The name of the flavour encodes information about the number of virtual CPUs, and the amount of memory and local disk space that the flavour provides; for example:
8c16m10d - CPUS: 8 - Memory: 16 GB - Local Disk: 10 GB
The standard set of flavours has 2GB of memory for each virtual CPU, and 10GB of local disk, i.e. 1c2m10d, 2c4m10d, and so on. The largest standard flavour is 8c16m10d, however special purpose flavours can be supported in certain cases.
For more information, see: Nirin Flavors and Charge Rates
Local Disk and Volume Storage: Local disk is only for the operating system install. This storage is not backed up anywhere, and only exists as long as the instance exists. It is strongly recommended that you deploy your instances using an automated process that permits you to rebuild an instance any time it is necessary, and to use persistent storage for anything that cannot be recreated easily.
Persistent storage is provided in the form of volume storage, and is backed onto a redundant and reliable storage backend. In addition, NCI provides tape based storage which is backed up across two remote sites - for more information on this, please contact the NCI Help Desk at help@nci.org.au.
Overview of storage types
Root disk | /dev/vda, automatically mounted on / | Operating system and applications |
|
|
Attached volumes aka Block Storage aka Cinder (Backed by Ceph) | /dev/vdb and subsequent, not automatically mounted | Persistent storage of large volumes of data |
|
|
NCI Global Data Filesystems | NFS exported (see Appendix C for more information) | Use of data sets across different NCI computing platforms, shared access to data from multiple projects, storage of very large data sets |
|
|
The NCI Global filesystems can be accessed via NFS exports. Please see NFS Export Policy for global file-systems to Nirin for more details.