Search results

I noticed a setting in BIOS recently (I believe it appeared in a recent update from SuperMicro) that read something like "expose CCX as NUMA" - I decided to turn it on... LSCPU: Indeed... NUMA "nodes" appear here (per CCX). I went through and adjusted all VM's... Enabled NUMA, assigned ~3-6...

Unfortunately, this isn't a valid option for modern enterprise environments. VBS and associated technologies are effectively required by contractual obligation for many businesses these days. There are security controls and configuration baselines that require it. My understanding is that...

I would really like to see an effort made to solve the nested virtualization problems so that we can use VBS in windows guests efficiently. I suspect the problem is not specific to Proxmox, rather KVM/QEMU specifically and any projects that use these. If anyone has any insight into what part of...

I've revisited VBS / credencial guard on Proxmox guests on our AMD cluster many times over the years. Have no seen much improvement on our cluster... When set to "host" CPU type, Windows 11/2025 will automatically attempt to enable VBS with nested extensions available, which causes high idle...

Fair question! My goal was to get the switch running software that is still "supported" by either a vendor or a project in some fashion, such that, vulnerabilities might get patched, and we would have a way to apply those patches. I don't think the Dell firmware is supported any longer. With...

It's not an "issue" there's just a lack of communication from within the guest to the host. VM's that can run the agent will produce better feedback here. Proxmox has to treat any memory that has been "touched" as used without further information about what it has been used for. I would say that...

Keep the VM settings simple/generic. Use IDE or SATA for the disk.

Another follow-up. All of the same problems returned a few weeks later. Totally random loss of communication on some networks between nodes. Ceph and Corosync seem to work fine on the E810+S532F's, but VLANS are a disaster. Coms will work randomly on some nodes and not on others. It's like...

EDIT: Never mind we tracked it down...

Same problem on one of our nodes today.... so far not having any luck with the solution that is working for others but I'm probably doing it wrong...

We had some growing pains trying to get E810 working but in the end, we did get them working. I don't think any of the issues we faced were actually related to Proxmox. Everything was related to low level hardware/port/operational/firmware settings at the cards and switches. FEC mode may need...

Following up... We disabled the LLDP agent and also disabled SR-IOV on both motherboards and cards since we don't use that feature anyway, then moved the networks back over to the new E810's. The Cluster and all networking has been working now for several weeks and through a normal monthly...

Hi Jsterr, We have disabled the LLDP agent on the E810's. Will make another attempt to move networks over later this week. Anyone have any thoughts on FEC? Should we be using it (RS mode) or None? I wonder if that could cause issues here.... Will follow-up later when we know more. Thanks

Hi jsterr, 15 character limit on bridge port name forced use of linux vlans for this config. We tested the 2-digit vlans directly bridged to the the interface with .XX and it made no differences. We also tested SDN and it made no difference, same problems. I do not believe our interfaces file...

We had to switch the FEC mode of the switch our E810's are connected to, to "RS" mode to get a link light.

Hello Proxmox Community! We are attempting to upgrade the core networking from 10Gb to 100Gb with a Dell S5232F at the helm for the cluster and a Dell S5248F for all of our network access switches to uplink to, intending to use 100Gb uplinks between the switches. The Switches have SONiC open...

I currently have ~400GB of DB/WAL space per 16TB drive on our cluster. A single NVME 2TB drive is servicing 4 X 16TB spinners in each server. It performed really well at first but then once they had been in service awhile the performance has suffered, especially in windows guests. I intend to...

Hi Superfish, The particular problems I have had with ceph write performance are mostly related to windows driver/cache stack issues. I get pretty good performance when running performance benchmarks directly on a pool. Your problem sounds to me like "sync-write" performance limitations of...

I would double check that your VM boot disks are indeed on that SSD. Ceph is software defined storage across a cluster of servers. I think it requires 3 servers minimum for testing / proof of concept, 4 minimum for low impact stuff like homelabs, 5+ for production clusters hosting commercial...

Slo VM boot speed is primarily driven by the disk performance backing the virtual boot disk. If your VM boot disks are all backed by those spinning drives, boot speeds of the VM's will be slow. I always back all VM boot disks with a ceph pool that has an "ssd only device class" rule. VM's boot...