Recent content by DamienDye

why are your drives formatted in 512e mode rather than native 4kn mode. has your ZFS pool also been configured to use 512 bytes too if so your by design wearing out your own disks because you haven't aligned them. have you 4k aligned your VM's with qm set <vmid> --args "-global...

just to follow up I have written a personal blog post about getting this working on the worst mini pc I have in the house. it has a script in the post if you would like to replicate my test enviroment. https://blogs.damiendye.uk/en/proxmox/flashstore-osd-on-proxmox/

Hi all, We’ve released a preview of FlashStore, croit’s Rust-based Ceph OSD for NVMe. It’s a drop-in replacement for the standard OSD. We think it could make a big difference to Ceph performance in hyper-converged Proxmox setups, so we’d really value feedback from Proxmox users. It’s packaged...

had a few people ask if they could put their own name on the VMs they sell so had a go and wrote it up, in my own time for something to do over the weekend. covers smbios1 and args for types 0/2/3/11 (serial from the vm name, sku from cores and memory), the seabios splash via -boot splash...

whats you benchmark results so we know what we are talking about

well am UK based and over provider connections I can get at least 2k mtu on private lines interconnects in DC are usually 9k in some its 12k even BT wholesale operates minijumbo but if there is a brain dead carrier blocking ICMP etc then yes you have no option but to lower the switch MTU for...

if you raise the raw mtu on the the current network you can then use 1500 mtu inside the vxlan tunnels. this should also mean that you don't have traffic issues and have to reduce the MTU on the nested instanced

actually there notes I have built over the last 25 years and when the notes where written windows 7 erah did remove it immediately

Yes this is possible and we have been providing assistance as a partner. if you want to reach out to us the site is in my bio.

don't forget to mark this as solved.

how have you created the mesh by hand or using the inbuilt SDN open fabric mesh note proxmox openfabric is the ISIS protocol not ospf because it's possible to turn the mesh into a ring while it's running what you need to do is move one cable add the node to the cluster and apply the SDN so...

am not sure this is the desired state you want a SDN network could be on any system supported as such you would be saying that having a system IP on a vTEP of vxnet should be supported too ? the SDN is meant to be an overlay network not the primary networks. you could just create a vlan on...

not much to go on here whats the logs say dmesg for example? what nic driver is it and whats it connected to in the system? are the nics on the same numa domain as the vCPU are you running EEE in the interface ? does it work faster without firewalling ?

like this you mean you can just enable it on the columns selector by hovering over a menu field and click the drop down

couple of questions 1) you open media vault is successfully connected and active on your tailnet ? 2) is tailscaled running on the omv system 3) do you see the tailscale0 interface on the omv with the tailscale ip once we have determined the basics are working we can move forward if any...