User Tools

Site Tools


server:nerdserv

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
server:nerdserv [2023/07/02 13:12]
ijon
server:nerdserv [2023/07/17 18:58] (current)
ijon
Line 7: Line 7:
 ===== What VMs can be on there? ===== ===== What VMs can be on there? =====
  
-TL;DR: Stuff that is beneficial to the crew, the station or is fulfilling our mission according to our bylaws (Satzungszweck). No mining, no torrenting, no illegal purposes, no wasting of ressources.+TL;DR: Stuff that is beneficial to the crew, the station or is fulfilling our mission according to our bylaws (Satzungszweck). No mining, no torrenting, no illegal purposes, no wasting of ressources, no tunnels or holes in Firewalls
  
 ==== Infrastructure ==== ==== Infrastructure ====
Line 20: Line 20:
   * Systems that create content for infoscreens like OpenMCT or Flowhub   * Systems that create content for infoscreens like OpenMCT or Flowhub
   * Collaboration stacks like Collabora, onlyoffice, hedgedoc ...   * Collaboration stacks like Collabora, onlyoffice, hedgedoc ...
 +  * PXE-Boot server
 +
 +Naming Scheme:
 +Hostname should give an indication what application(s) are running on the machine
  
 ==== Personal ==== ==== Personal ====
Line 34: Line 38:
   * "I need this to run overnight and dont want to leave my notebook on board"   * "I need this to run overnight and dont want to leave my notebook on board"
   * "I need a development machine for my (educational / self-teaching) project"​   * "I need a development machine for my (educational / self-teaching) project"​
 +
 +Naming Scheme:
 +Hostname starts with the ldap-name of the member, followed by a - and then some indication of the purpose of the machine
  
 ==== Ressource wasting ==== ==== Ressource wasting ====
Line 50: Line 57:
 Edgecases: Continous Integration Pipelines can, depending on how often they run and how large the project is, really eat up all provided ressources quickly. In general, this may not be a good fit for nerdserv. Talk to admax@ before you deploy systems that will eat up 100% of the provided cores over many hours. Edgecases: Continous Integration Pipelines can, depending on how often they run and how large the project is, really eat up all provided ressources quickly. In general, this may not be a good fit for nerdserv. Talk to admax@ before you deploy systems that will eat up 100% of the provided cores over many hours.
  
-===== How to get a VM? =====+===== How to get a LXC or VM? =====
  
-There are two ways to get a VM+There are two ways to get an LXC or a VM
  
 ==== Infrastructure as Code ==== ==== Infrastructure as Code ====
  
-If you are able to use Github and the task of "​creating a pull request"​ does not scare you, then we'd like to ask you, to specify your VM yourself. ​+If you are able to use Github and the task of "​creating a pull request"​ does not scare you, then we'd like to ask you, to specify your VM (or LXC) yourself. ​
  
 To do this, clone  To do this, clone 
 https://​github.com/​c-base/​nerdserv https://​github.com/​c-base/​nerdserv
  
-Inside the Folder /proxmox, you can duplicate any of the machine files, like homeassistant.tf,​ rename the file and edit the contents so it matches your requirements. Dont forget to include your ssh pub cey.+Inside the Folder /proxmox, you can duplicate any of the machine files, like homeassistant.tf,​ rename the file and edit the contents so it matches your requirements. Dont forget to include your ssh pub cey in the locals.tf file. After creation, your username for this SSH-Key initially is "​debian"​. Please change it.  
 + 
 +//This Tutorial is in so far incomplete, as it has been discovered that the OS needs to be specified and the SSH key of the user wanting the machine needs to be added here as well. How this works can be edited by anyone who is reading this.// ​
  
 Then commit and create a pull request. After the pull request is merged, terraform will perform its magic and the VM (or LXC Container) will pop into existence. ​ Then commit and create a pull request. After the pull request is merged, terraform will perform its magic and the VM (or LXC Container) will pop into existence. ​
Line 67: Line 76:
 ==== Manual ==== ==== Manual ====
  
-Please specify number of Cores, amount of RAM, usage scenario, Disk space, hostname, ssh-pub-cey,​ end-date (if possible) and request your VM through writing an email to admax@+Please specify ​if you need a VM or LXC, the OS you like to have on there, the number of Cores, amount of RAM, usage scenario, Disk space, hostname, ssh-pub-cey,​ end-date (if possible) and request your VM through writing an email to admax@ 
 + 
 +===== System Specification ===== 
 + 
 +==== Hardware ==== 
 + 
 +Nerdserv runs on a Super Micro board called a2sdi-16c-hln4f. This board has a Intel Atom Processor C3955 with 16 Cores.  
 +The board is equipped with 128GB of ECC RDIMM RAM, which can be extended to 256GB of RAM, if two RAM Modules (RDIMM ECC DDR4 2400 64GB) are bought and added 
 + 
 +The OS is on two SSDs that are in a RAID1 configuration. This is exclusivly for the Hypervisor. 
 + 
 +VMs get their disk space from a M.2 NVME Disk that has 2TB capacity.  
 + 
 +==== Software ==== 
 + 
 +As Hypervisor, Proxmox is used.  
 + 
 +==== Storage ==== 
 + 
 +If you require signifcant amounts of storage ("more then a couple of GB"), please use the MFS as (mounted) storage 
 + 
 +==== Backups ==== 
 + 
 +Maybe a snapshot from your VM is regularly created and saved on the MFS. Do not rely on this. 
server/nerdserv.1688296342.txt.gz · Last modified: 2023/07/02 13:12 by ijon