XCP-ng
    • Categories
    • Recent
    • Tags
    • Popular
    • Users
    • Groups
    • Register
    • Login

    Some guidelines for sizing a XO server ?

    Scheduled Pinned Locked Moved Xen Orchestra
    5 Posts 2 Posters 511 Views 2 Watching
    Loading More Posts
    • Oldest to Newest
    • Newest to Oldest
    • Most Votes
    Reply
    • Reply as topic
    Log in to reply
    This topic has been deleted. Only users with topic management privileges can see it.
    • HolgiBH Offline
      HolgiB
      last edited by

      Hey there,

      we are using XO build from sources in a Ubuntu VM. Mainly to have scheduled backups of important VMs but we are slowly adopting to use XO instead of XCP-ng Center (which is depricated anyway) for administration purposes.

      Are there any guidelines / rule of thumb for the sizing of the VM in regard of CPU cores / RAM ?
      I mean like 8 cores and 16 GB for managing 400 VM+ ?

      TIA,
      Holger

      nick.lloydN 1 Reply Last reply Reply Quote 1
      • nick.lloydN Offline
        nick.lloyd @HolgiB
        last edited by

        @HolgiB XO doesn't require a huge amount of resources. There are limits of having 64 hosts in one pool, but that's the only one I'm aware of.

        I'd start with 2c 4gb RAM for anything relatively small (<100 VMs). All of the processing is done on the pool master (unlike VMware), so it's unlikely you'll run into performance hits unless you plan on using the backup feature heavily. It's also pretty easy to change afterwards too. Just select the XO VM, click on the field you want to change, enter desired values, and reboot!

        HolgiBH 1 Reply Last reply Reply Quote 0
        • HolgiBH Offline
          HolgiB @nick.lloyd
          last edited by

          @Nick-085 Thanks for answering. We have been using XO on a rather "beefy" VM with 6 cores and 16 GB RAM. Lately we added a lot of backup jobs to the host and run into some issues of backups failing because XO reported that the backup remote is partially not availabe. After increasing RAM the issue seemed to have moved away. I guess I will have a close eye on this but not invest too much time if more RAM solves the problem.

          nick.lloydN 1 Reply Last reply Reply Quote 0
          • nick.lloydN Offline
            nick.lloyd @HolgiB
            last edited by

            @HolgiB Perfect! The other thought I had would be adding more XO VMs to distribute the backup load, or add a Backup Proxy (assuming you used this script, you can deploy a proxy in XO by running the script again and selecting the "deploy proxy" option).

            HolgiBH 1 Reply Last reply Reply Quote 0
            • HolgiBH Offline
              HolgiB @nick.lloyd
              last edited by

              @Nick-085 Thanks for the hint mate ! I wasn't aware that this script also installs a XO proxy. In the mean time we already have something like backup proxies in place because we splitted backup jobs to two XO instances in the first place since we wanted to separate test environments from "production" (so to speak). A third instance is used mainly for centralized administration (we are only slowly moving away from using "Xencenter" since we are long time Citrix Xenserver users). I guess old habbits die slowly. In the end I guess by coincidence we already build something similar to what a central XO instance plus two XO proxies 🙂

              1 Reply Last reply Reply Quote 1
              • First post
                Last post