Recent content by bbgeek17

Any NFS will work seamlessly with Proxmox. The protocol abstracts all complexity away. However, with NFS come the latency issues you are alluding to. The primary way to get low latency response is to remove as much layering as possible, i.e. go directly to raw storage. And/or use storage that...

Thats the nature of the game, Proxmox uses LVM in most basic installation. In other cases it could be ZFS, or NFS, etc. The OS you install will use LVM/ext/zfs/NTFS/etc. Dont overthink it. The most natural way to install is with LVM Thin. Its also the most flexible down the road. When you get...

The massive farms run by governments, universities, public and private corporations wont notice your absence. Its up to you! Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

Any cyclic dependency can lead to unpredictable issues down the road. OPsense depends on Proxmox to function, Proxmox depends on Opsense. Granted NTP may not be critical (except some cases). If Proxmox can access outbound internet without relying on OPsense, then why not go directly to the...

You need to quote space containing strings, or better yet - don't use spaces in exports. Good luck. Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

@bomzh is correct, ZFS is not a cluster aware filesystem, you will have data corruption if you attempt to use it as such. You may be able to use ZFS over ISCSI with your setup. Good luck. Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

fio is the industry standard way to test performance of a disk. Good luck. Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

There is nothing you can do in the software about NO-CARRIER. It means no physical/data-link layer. Move the cable around to other NICs, you have 3 others. May be you are the unlucky owner of a card based in Realtek chipset (search forums). May be something even more exotic. Try to boot into a...

The file you showed in your screenshot : vi /etc/network/interfaces Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

Just edit the file and change eth0 to eth1 eno1, which ever one works for you. No, I dont know how it was installed, what has changed (kernel update?) or what you may have changed. Its an easy fix. Blockbridge : Ultra low latency all-NVME shared storage for Proxmox -...

the vmbr0 is configured to use eth0, which does not exist Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

If you list the exact steps you took, add your configuration file (using CODE tags or similar), add output commands of your steps, someone might be able to help. You can also always reach out to your Storage vendor (Dell) and ask for assistance in configuring this particular model for multipath...

Then your multipath is not configured properly. Is that 2x32 devices, or 1 with 4 paths? Either way, you need to review your configuration. Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox

You said you configured multipath, this path ^ does not look like a correct one for a multipath owned disk. What do the following commands show: lsscsi (may need to be installed) lsblk multipath -ll If you use multipath, all volume manager manipulations must be against the DM multipath device...

In best case scenario the configuration can be driven completely from PVE GUI. In more complex environments you may need to drop to shell. Some helpful links are: https://pve.proxmox.com/pve-docs/pve-admin-guide.html#storage_open_iscsi https://pve.proxmox.com/wiki/Storage:_iSCSI...