Difference between revisions of "NikXOLocalService"
From PDP/Grid Wiki
Jump to navigationJump to search| Line 1: | Line 1: | ||
= NikXO Local Virtual Machine Service = | = NikXO Local Virtual Machine Service = | ||
| − | The NikXO Local (NXOL) service provides stand-alone virtual machines as Infrastructure-as-a-Service (IaaS) to Nikhef users, using a self-service | + | The NikXO Local (NXOL) service provides stand-alone virtual machines as Infrastructure-as-a-Service (IaaS) to Nikhef users, using a self-service auto-provisioning portal based on XenOrchestra and a dedicated network to which access is liberally open from within Nikhef, but that is (by default and by design) closed to the outside world. It allows Nikhef users to try out stuff on a virtual machine that they can and should manage by themselves, but whose management they ''may'' share with everyone at Nikhef. |
=== Important notice === | === Important notice === | ||
| Line 19: | Line 19: | ||
== What can you expect == | == What can you expect == | ||
| − | The NXOL service provides one-off VMs that are stand-alone, but can use the standard Nikhef account management (LDAP) functionality if you want to. You cannot do NFS traffic from here to other Nikhef systems, although CIFS (SMB, Windows) access may work if you try with a samba | + | The NXOL service provides one-off VMs that are stand-alone, but can use the standard Nikhef account management (LDAP) functionality if you want to. You cannot do NFS traffic from here to other Nikhef systems, although CIFS (SMB, Windows) access may work if you try with a samba client on your machine. |
| − | + | and can be renewed, but since you share the network with all other Nikhef users of the NXOL service, please be kind provisioning | |
| Line 33: | Line 33: | ||
You can create VMs with a few cores (1-4), reasonable memory (2-4GByte or so) and some disk (say, 16-30GByte) without trouble and without exhausting the shared pool of resources assigned to all NXOL users together. | You can create VMs with a few cores (1-4), reasonable memory (2-4GByte or so) and some disk (say, 16-30GByte) without trouble and without exhausting the shared pool of resources assigned to all NXOL users together. | ||
| − | + | randomly pick an address. If you do, we will have to abandon the provisioning service for everyone. Do be kind :) | |
| Line 41: | Line 41: | ||
== Stopping VMs == | == Stopping VMs == | ||
| − | + | If you | |
== CentOS7 PXE provisioning service == | == CentOS7 PXE provisioning service == | ||
| Line 47: | Line 47: | ||
You want to (re)install your VM and are happy with CentOS7 64-bit (CentOS is, like Scientific Linux, a RHEL clone), and want to use your local Nikhef account there to login and gain root access, as well as have some basic settings done right? Use the PXE provisioning service by (re)starting your VM using PXE network boot and DHCP. | You want to (re)install your VM and are happy with CentOS7 64-bit (CentOS is, like Scientific Linux, a RHEL clone), and want to use your local Nikhef account there to login and gain root access, as well as have some basic settings done right? Use the PXE provisioning service by (re)starting your VM using PXE network boot and DHCP. | ||
| − | Before you continue: * register your (password-protected!) SSH keys in the Nikhef IdM system by going to '''https://sso.nikhef.nl/chsh/''', copy your | + | Before you continue: * register your (password-protected!) SSH keys in the Nikhef IdM system by going to '''https://sso.nikhef.nl/chsh/''', copy your SSH public keys (<code>ssh-rsa AAAAB3NzaC1y...</code>) into the box (you may enter more then one) and then confirming the account update. |
Then, DHCP-boot the VM and wait for the PXE boot prompt to appear. The initial menu looks like this | Then, DHCP-boot the VM and wait for the PXE boot prompt to appear. The initial menu looks like this | ||
| Line 80: | Line 80: | ||
| − | + | this effective'''. Now your machine will boot quickly next time. | |
