gurubert's latest activity

  • gurubert
    gurubert reacted to Red Squirrel's post in the thread How to access VM's from anywhere with Like Like.
    I would setup a VPN server like Wireguard or OpenVPN. Once you have that setup you VPN in and then can access everything on your network. I have mine setup in a way where I need to authenticate myself first on a hidden web page which unblocks...
  • gurubert
    gurubert reacted to jtru's post in the thread Quake III Arena game server launch with Like Like.
    From PVE's perspective, q3ded is an application for a Linux host as any other. So you just set up an LXC container or QEMU VM with a GNU/Linux distribution of your choice in it on PVE, make sure it's properly available on the local network, and...
  • gurubert
    gurubert reacted to tchaikov's post in the thread ceph / cephfs feature level mismatch with Like Like.
    the documentation might benefit from a note that using cephfs will (as of now) prevent use of some current ceph features. agreed. the kernel cephfs does not advertise the reef features. because the pg-upmap-primary needs the cooperation of...
  • gurubert
    the VLAN SDN zone already does the clusterwide syncing heavyweight lifting. nothing stops you from creating a script that polls VLAN list from the external source and uses native CLI tooling to add VLAN entries. Then, cron is your friend :)
  • gurubert
    gurubert reacted to Johannes S's post in the thread Überwachen ob ein VM lebt. with Like Like.
    An und für dich eine gute Empfehlung, aber der Op will ja Monitoring ohne Monitoring-Software. Diesen Wunsch nach einer Quadratur des Kreises wird auch uptimekarma nicht erfüllen können
  • gurubert
    gurubert replied to the thread ProxMox SDN.
    The cluster should come back.
  • gurubert
    gurubert replied to the thread ProxMox SDN.
    All VMs will stop working as the Ceph cluster will stop working. The Proxmox cluster also will stop working.
  • gurubert
    Thanks for the logs. So to me it seems like there was a network communication issue where the nodes 1 and 2 couldn't communicate with each other over the Ceph interfaces. Corosync seems fine according to the log. My reason to believe this: Log of...
  • gurubert
    Thanks, that was my conclusion as well. As a next step, I will examine the network more closely. There are plans to expand the cluster by one or two hosts in the near future. Consequently, this is unfortunately not an option.
  • gurubert
    There is the vmfs-tools package available on Debian. But this is out of the support scope of Proxmox.
  • gurubert
    Hi @FBFM, there is no official PVE method of mounting VMFS on a locally attached SAN LUN. There are several migration methods described here: https://pve.proxmox.com/wiki/Migrate_to_Proxmox_VE , including an automatic import from running ESXi...
  • gurubert
    Problem solved. I realized that due to secure-boot drbd 9 was rejected and the system was using drbd 8. I had to re-install everything, properly sign drbd 9 and use mokutil to enroll. after that. After that I had to re-start from scratch with the...
  • gurubert
    gurubert reacted to UdoB's post in the thread ProxMox SDN with Like Like.
    I do not use SDN, so I can not help. For "risky" experiments (originally with Ceph, in my case) I've setup a virtual PVE cluster. Inside that cluster one can run dangerous experiments without the fear to destroy something important - just create...
  • gurubert
    gurubert reacted to jdancer's post in the thread CEPH & ZFS Using a PERC H755n with Like Like.
    As you know, ZFS & Ceph really don't play nice with RAID controllers like Dell PERCs. While it's true you change the PERC from RAID mode to IT-mode, I've had issues in the past with the megaraid_sas driver. That's why I switch out Dell PERCs for...
  • gurubert
    Isn't the issue here that the client process requesting the data from the OSDs needs to know where in the CRUSH topology it runs? In a hyperconverged cluster with only three nodes this may be a very good optimization, because there all data is...