Search results for query: affinity

  1. I

    Higher CPU utilization on i9-13900H vs i9-12900H for same VM – Proxmox 9.1.5

    ...no I/O wait on either node. Host CPU (us+sy) is minimal (~4-5%) on both. P-CORE PINNING TEST Pinning a VM to P-cores only (qm set <vmid> -affinity 0-11) on node-02 dropped CPU from ~24% to ~17%. This confirms E-core scheduling accounts for part of the difference, but there’s still a ~6% gap...
  2. G

    Schlechte NUMA-Konfigurierbarkeit für KVM-Gäste.

    ...633+9.91%1.099× Wir bekommen also knappe 10% Leistungsunterschied. Feature-Request Cool wäre es, wenn man den numaX-Zeilen auch eine Affinity mitgeben könnte, die vor der Ausführung eines Hook-Scripts gesetzt würden: numa0: cpus=0-3,hostnodes=0,affinity=0-31,memory=81920,policy=bind...
  3. H

    [SOLVED] CPU Arguments cuts passthroughed NVMe performance by 4 times

    ...args by myself? Here is my 100.conf file, the NVMe I passthrough is the hostpci2. It is not in any other IOMMU group with anything other. affinity: 0-11 args: -cpu...
  4. G

    Schlechte NUMA-Konfigurierbarkeit für KVM-Gäste.

    ...höhere Performance. MSSQL hat NUMA support. Aber in Proxmox unter der Konfiguration, wenn man nur NUMA (numa: 1) aktiviert (kein affinity bei < vCPUs als logische CPUs einer NUMA-Node, etc.) sehe ich nicht, wie das sein kann. Tests wie oben sind sehr aufwändig und einer mit MSSQL wäre für...
  5. M

    Restore VM data after HA / replication sync issue – node failure scenario

    ...on B to start then the replication job will have switched automatically from B -> A. The VM running on B will stay there barring any affinity rules [2]. If your VM state on node A is inconsistent/bad you can force a migration to node B, this worked on my test setup: * Make sure node A is...
  6. G

    Schlechte NUMA-Konfigurierbarkeit für KVM-Gäste.

    ...das mal mit Benchmarks für Speicherzugriff (leerer Host, getestet mit likwid-bench, VM mit 8 vCPUs und 64GiB RAM): Gedanken numa_balancingaffinitynumahostnodesload (MB/s)stream_mem (MB/s) Die Worst-Case-Baseline (Memory und CPUs auf unterschiedlichen NUMA-Nodes.)00-3111108798.89129484.39...
  7. G

    Schlechte NUMA-Konfigurierbarkeit für KVM-Gäste.

    ...mit jeweils 2xEPYC 9375F, also 2 NUMA-Nodes pro Server, wenn wir die einzelnen CCDs ignorieren. Relevante Konfigurationsparameter sind: affinity, numa und numa0. Wenn ich also eine VM habe, die in eine NUMA-Node (RAM sowie weniger als NUMA-Node CPU Kerne -2 oder so) passt ist das an für sich...
  8. guletz

    Proposed Hardware Design – Proxmox VE Compatibility & Shortfalls

    Hi, It will be better to have only one CPU with so many cores as you can afford! With 2 socket you will need to play with affinity. Good luck / Bafta !
  9. C

    Adding HA resources to an existing HA affinity rule and outdated manpages

    ...$VMID --group $groupname to be able to assign a VM to what then was a "rule" (HA Groups). With PVE 9 HA Groups have been removed in favor of affinity rules which do offer more functionality - But I am wondering how I can add a HA resource (e.g. vm:123) to an already existing rule via the CLI...
  10. F

    Have trouble with B50 passthrough in PVE9 6.17 kernel.

    ...107.conf 108.conf 110.conf 113.conf 200.conf 301.conf 302.conf 500.conf root@Flanker9900T:~# cat /etc/pve/qemu-server/113.conf affinity: 0-15 agent: 1 balloon: 4096 bios: ovmf boot: order=scsi0 cores: 8 cpu: host efidisk0...
  11. cwt

    Frage zu 4 Sockets a 4 Cores vs 1 Socket a 16 Cores

    ...Wenn wirklich mehrere CPUs vorhanden sind, sollte man im Bereich einer CPU bleiben pro VM. Ansonsten gibt es downsides bei NUMA. Mittels CPU affinity kann man vCores auf eine CPU festnageln. Selbst bei großen single socket CPUs lohnt sich eine etwas genauere Planung der vCPU Verteilung...
  12. R

    [SOLVED] Verständnis Frage HA

    ...Darauf läuft nur ein Domaincontroller und ein PiHole. Die VM's der beiden großen Hosts habe ich in HA eingebunden und für jede VM eine "Affinity Rule" gebaut, dass diese nur auf Host 1 und 2 laufen dürfen. -> Hier die Frage, muss ich das? Weil mangels Storage können die VM's ja auch nicht...
  13. cwt

    Terminalserver unter PVE langsam

    ...und 4GB RAM (wenn er auch nur als reiner DC läuft). 4 vCPUs für den TS ist dann wiederum wenig. Du könntest in der conf der TS VM die CPU affinity festlegen (bspw. 0-9) und NUMA auf off. Das würde cross NUMA bzw. Verteilung auf 2 physische CPUs verhindern. Wie sieht denn die weitere conf...
  14. VictorSTS

    HA Migration issues when VM has local storage

    ...[1]. As HA is right now, it's fully admin responsibility to check the systems configuration (i.e. storage type or limit HA group/affinity to the nodes that do have the VM's disk(s) ). Of course it would be better if PVE did some checks to prevent human errors, but they are not implemented...
  15. H

    RTX 5060 Ti VFIO passthrough on AMD X570 consumer motherboard fails – Firmware 1:1 IOMMU issue

    ...Indicator = Disabled Resizable BAR = Enabled. I think the Kernel DMA Protection Indicator is the key to this. You can also try set Pre-boot DMA Protection to disable. One issue that I'm having now is CPU usage is very high when gaming and have very low fps. Already tried CPU affinity. Any Ideas?
  16. T

    In Proxmox VE (PVE), the memory allocation is uneven when checking with numactl --hardware.

    ...allocated a large amount of memory (180GB). Even if I shut down this VM and restart it, after some time, running the numactl --hardware command to check the physical machine's memory distribution causes the entire system to lag. However, I still want to manually configure the CPU affinity...
  17. D

    proxmox-cpu-affinity service - version 0.0.9

    Well :) this line - is the online line in Proxmox to set task affinity. https://github.com/proxmox/qemu-server/blob/e781d9713fb3d45a523fe14db6cd99ac60bbfbb5/src/PVE/QemuServer.pm#L3202 Even with Numa (where you have multiple CPUs that have direkt and not direct memory) the user (that's you!)...
  18. fstrankowski

    proxmox-cpu-affinity service - version 0.0.9

    ...in having the system autobalancing the load, especially for a high amount of vms should be considered aswell. Personally, i would not set affinity for each and every VM because i'd lose lots of dynamic behaviour and loadbalancing and rather decide on a case-by-case basis if the VM requires a...
  19. D

    proxmox-cpu-affinity service - version 0.0.9

    We now have an SVG exporter. VMs are automatically assigned. proxmox-cpu-affiny tries to avoid bad combos. root@proxmox:~# proxmox-cpu-affinity status svg --affinity -o heatmap-affinity.svg This is still not production ready but it runs now on all of my servers.
  20. D

    proxmox-cpu-affinity service - version 0.0.9

    Hello Proxmoxer, I created a tool to automatically set the CPU affinity (at runtime) for VMs. https://github.com/egandro/proxmox-cpu-affinity This picks the best combination of cores. As mentioned by this project: https://github.com/nviennot/core-to-core-latency. Here an example. CPU 1...