d-bis/Sankofa
4
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
Files
main
Sankofa/docs/infrastructure/PROXMOX_VE_HARDWARE_BOM.md
defiQUG ee551e1c0b Update Proxmox VM specifications and optimize deployment configurations 
- Revised CPU and memory specifications for various VMs, moving high-resource workloads from ML110-01 to R630-01 to balance resource allocation.
- Updated deployment YAML files to reflect changes in node assignments, CPU counts, and storage types, transitioning to Ceph storage for improved performance.
- Enhanced documentation to clarify resource usage and deployment strategies, ensuring efficient utilization of available hardware.
2025-12-13 04:46:50 -08:00

28 KiB
Raw Permalink Blame History

Sankofa Phoenix: Proxmox VE Hardware Bill of Materials (BOM)

Date

2025-01-XX

Overview

This document provides a comprehensive inventory of all Proxmox VE hardware in the Sankofa Phoenix infrastructure, including current hardware specifications, available hardware, Proxmox VE compatibility notes, and deployment recommendations.

Current Hardware Inventory

Summary Statistics

Category Count Total RAM Total CPU Cores (Validated) GPU Systems
Total Systems 16 2,304 GB 34+ cores (validated) 2
HPE ML110 Gen9 1 256 GB 6 cores 0
Dell R630 (High Memory) 1 768 GB 52 cores (104 threads) 0
Dell R630 (Standard) 12 1,536 GB ~260-390 cores (est.) 0
Dell Precision 7920 2 192 GB ~32-64 cores (est.) 2

Note:

  • ✅ HPE ML110 Gen9: Validated (6 cores)
  • ✅ Dell R630 (High Memory): Validated (52 cores, 104 threads)
  • ⏳ Dell R630 (Standard): Estimates based on typical configurations
  • ⏳ Dell Precision 7920: Estimates based on typical configurations

Detailed Hardware Specifications

1. HPE ML110 Gen9

System ID: PVE-HOST-001
Hostname: ml110-01
IP Address: 192.168.11.10
Status: Active
Proxmox VE Version: 9.1.1 (pve-manager/9.1.1/42db4a6cf33dac83)
Kernel: 6.17.2-1-pve
Cluster: [To be determined]

Hardware Specifications

Chassis:

  • Manufacturer: Hewlett Packard Enterprise (HPE)
  • Model: ProLiant ML110 Gen9
  • Form Factor: Tower Server
  • Rack Mountable: Optional (with rack kit)
  • Serial Number: [Not available via DMI]

Processor:

  • CPU Model: Intel Xeon E5-2603 v3 @ 1.60GHz
  • CPU Count: 1 processor (single socket)
  • CPU Cores: 6 cores
  • CPU Threads: 6 threads (no hyperthreading)
  • CPU Architecture: x86_64
  • CPU Speed: 1.60 GHz (Base), 1.20-4.00 GHz (Range)
  • CPU Family: Xeon (Haswell-EP)
  • CPU Stepping: 2
  • Virtualization: Intel VT-x (VMX) supported
  • Cache:
    • L1d: 192 KiB (6 instances)
    • L1i: 192 KiB (6 instances)
    • L2: 1.5 MiB (6 instances)
    • L3: 15 MiB (1 instance)

Memory:

  • Total RAM: 256 GB (251 GiB usable)
  • RAM Type: DDR4 ECC LRDIMM (Load-Reduced DIMM)
  • Memory Modules: 8x 32 GB modules
  • Memory Speed: 2133 MT/s (configured at 1600 MT/s)
  • Memory Configuration: Multi-bit ECC
  • Memory Slots: [To be determined - likely 8-16 slots]
  • Available Memory: ~244 GB (for VMs)

Storage:

  • Storage Controller: Intel C610/X99 series chipset 6-Port SATA Controller (AHCI mode)
  • Storage Disks:
    • 2x Seagate ST1000DM003-1ER162 (1TB SATA HDD)
    • sda: 931.5 GB (primary, with Proxmox installation)
    • sdb: 931.5 GB (secondary, used for Ceph OSD)
  • Storage Configuration:
    • Primary disk (sda): LVM with Proxmox VE installation
      • pve-swap: 8 GB
      • pve-root: 96 GB
      • pve-data: 794.3 GB (for VMs)
    • Secondary disk (sdb): Ceph OSD block device
  • RAID Configuration: Software-based (LVM, Ceph)
  • Storage Options: SATA AHCI (no hardware RAID controller detected)

Network:

  • Network Adapters: 2x Broadcom NetXtreme BCM5717 Gigabit Ethernet PCIe
  • Network Ports: 2x 1GbE ports
    • nic0 (enp2s0f0): Active, connected to vmbr0 bridge
    • nic1 (enp2s0f1): Available (not configured)
  • Network Bridges: vmbr0 (192.168.11.10/24)
  • MAC Addresses:
    • nic0: 1c:98:ec:52:43:c8
    • nic1: 1c:98:ec:52:43:c9
  • Additional NICs: Supports PCIe expansion cards

Power:

  • Power Supply: [To be determined]
  • Power Rating: [To be determined]
  • Power Consumption: [To be determined]

Proxmox VE Compatibility:

  • ✅ Fully Compatible - HPE ML110 Gen9 is fully supported by Proxmox VE
  • Installed Proxmox VE Version: 9.1.1 (latest stable)
  • Virtualization Support: Intel VT-x (VMX) enabled and working
  • Storage:
    • Local LVM storage configured
    • Ceph OSD configured on secondary disk
    • CephFS mounted at /mnt/pve/ceph-fs (384 GB available)
  • Network: Standard network bridges configured (vmbr0)
  • Boot Mode: UEFI (EFI boot mode)

Current Configuration:

  • Proxmox VE: Installed and operational
  • Storage Pools:
    • local-lvm: 794.3 GB available (LVM thin pool)
    • ceph-fs: 384 GB available (Ceph filesystem)
  • Network: vmbr0 bridge configured with static IP (192.168.11.10/24)
  • VMs Running: Multiple VMs configured (VMIDs: 136, 139, 141, 142, 145, 146, 150, 151)
  • Ceph: Ceph OSD configured on secondary disk

Deployment Notes:

  • ✅ Currently Active - System is operational and hosting VMs
  • Suitable for development/testing workloads
  • Can serve as Proxmox VE cluster node
  • Recommended for low-to-medium workload VMs
  • Consider for backup/storage node
  • CPU Limitation: 6 cores may limit concurrent VM performance
  • Storage: Using software-based storage (LVM, Ceph) - no hardware RAID
  • Network: 1GbE network - consider upgrade to 10GbE for better performance

2. Dell PowerEdge R630 (High Memory)

System ID: PVE-HOST-002
Hostname: r630-01
IP Address: 192.168.11.11
Status: Active
Proxmox VE Version: 9.1.1 (pve-manager/9.1.1/42db4a6cf33dac83)
Kernel: 6.17.2-1-pve
Serial Number: HNQ3FB2
UUID: 4c4c4544-004e-5110-8033-c8c04f464232
Cluster: [To be determined]

Hardware Specifications

Chassis:

  • Manufacturer: Dell Inc.
  • Model: PowerEdge R630
  • Form Factor: 1U Rack Server
  • Rack Mountable: Yes
  • Serial Number: HNQ3FB2

Processor:

  • CPU Model: Intel Xeon E5-2660 v4 @ 2.00GHz
  • CPU Count: 2 processors (dual socket)
  • CPU Cores: 26 cores per processor (52 total cores)
  • CPU Threads: 26 threads per processor (104 total threads with hyperthreading)
  • CPU Architecture: x86_64 (Broadwell-EP)
  • CPU Speed: 2.00 GHz (Base), 1.20-3.20 GHz (Range, Turbo up to 3.20 GHz)
  • CPU Stepping: 1
  • Virtualization: Intel VT-x (VMX), VT-d supported
  • NUMA: 2 NUMA nodes (one per CPU socket)
  • Cache:
    • L1d: 1,664 KiB (52 instances)
    • L1i: 1,664 KiB (52 instances)
    • L2: 13 MiB (52 instances)
    • L3: 130 MiB (2 instances, 65 MiB per socket)

Memory:

  • Total RAM: 768 GB (755 GiB usable, 792 GB total including system overhead)
  • RAM Type: DDR4 ECC LRDIMM (Load-Reduced DIMM)
  • Memory Modules: 12x 64 GB modules
    • Part Number: M386A8K40BM1-CRC (Samsung)
    • Speed: 2400 MT/s
    • Type: Synchronous Registered (Buffered) LRDIMM
    • Error Correction: Multi-bit ECC
  • Memory Slots: 24 DIMM slots (12 per CPU socket)
  • Memory Configuration: 6 modules per CPU socket (A1-A6 populated, A7-A12 empty)
  • Available Memory: ~744 GB (for VMs)
  • Maximum Memory: Up to 1.5 TB (with additional LRDIMMs)

Storage:

  • Storage Controller: Dell PERC H730 Mini (LSI MegaRAID SAS-3 3108 [Invader])
  • Storage Disks:
    • 2x 300GB drives
    • sda: Seagate ST9300653SS (279.4 GB, primary with Proxmox installation)
    • sdb: HUC106030CSS600 (279.4 GB, secondary used for Ceph OSD)
  • Storage Configuration:
    • Primary disk (sda): LVM with Proxmox VE installation
      • pve-swap: 8 GB
      • pve-root: 79.6 GB
      • pve-data: 171.3 GB (for VMs)
    • Secondary disk (sdb): Ceph OSD block device
  • RAID Configuration: Hardware RAID controller (PERC H730 Mini)
  • Storage Bays: 10x 2.5" hot-swappable drive bays
  • Storage Options: SATA, SAS, NVMe (with riser card)

Network:

  • Network Adapters: 4x Broadcom NetXtreme II BCM57800 1/10 Gigabit Ethernet
  • Network Ports: 4x 10GbE ports (1/10 Gigabit capable)
    • nic0 (enp1s0f0): Available (not configured)
    • nic1 (enp1s0f1): Available (not configured)
    • nic2 (enp1s0f2): Active, connected to vmbr0 bridge
    • nic3 (enp1s0f3): Available (not configured)
  • Network Bridges: vmbr0 (192.168.11.11/24)
  • MAC Addresses:
    • nic0: c8:1f:66:d2:c5:97
    • nic1: c8:1f:66:d2:c5:99
    • nic2: c8:1f:66:d2:c5:9b (active)
    • nic3: c8:1f:66:d2:c5:9d
  • Network Capabilities: 10GbE capable (currently configured for 1GbE)
  • Additional NICs: Supports PCIe expansion cards for 25GbE/100GbE

Power:

  • Power Supply: Dual redundant power supplies (typical for R630)
  • Power Rating: [To be determined - typically 495W, 750W, or 1100W]
  • Power Consumption: [To be determined]

Proxmox VE Compatibility:

  • ✅ Fully Compatible - Dell R630 is fully supported by Proxmox VE
  • Installed Proxmox VE Version: 9.1.1 (latest stable)
  • Virtualization Support: Intel VT-x (VMX), VT-d enabled and working
  • Storage:
    • Local LVM storage configured
    • Ceph OSD configured on secondary disk
    • Hardware RAID controller (PERC H730) available
  • Network: Standard network bridges configured (vmbr0), 10GbE capable
  • Boot Mode: UEFI (EFI boot mode)
  • High Memory: Excellent for memory-intensive workloads

Current Configuration:

  • Proxmox VE: Installed and operational
  • Storage Pools:
    • local-lvm: 171.3 GB available (LVM thin pool)
    • Ceph OSD: Configured on secondary disk
  • Network: vmbr0 bridge configured with static IP (192.168.11.11/24)
  • VMs Running: Multiple VMs configured (VMIDs: 101, 104, 134, 137, 138, 144, 148)
  • Ceph: Ceph OSD configured on secondary disk
  • CPU Utilization: 104 logical CPUs available (52 cores × 2 sockets with hyperthreading)
  • Memory Utilization: ~744 GB available for VMs

Deployment Notes:

  • ✅ Currently Active - System is operational and hosting VMs
  • Ideal for high-memory workloads (databases, in-memory caches)
  • Excellent for Proxmox VE cluster node
  • Can host many VMs with high memory requirements
  • Consider for primary compute node in cluster
  • CPU: 52 cores (104 threads) provides excellent compute capacity
  • Memory: 768 GB provides excellent capacity for memory-intensive workloads
  • Storage: Hardware RAID controller available (PERC H730 Mini)
  • Network: 10GbE capable - consider configuring additional ports for better performance

3-14. Dell PowerEdge R630 (Standard Configuration)

System IDs: PVE-HOST-003 through PVE-HOST-014
Quantity: 12 systems
Status: Active
Proxmox VE Version: [To be determined]
Cluster: [To be determined]

Hardware Specifications

Chassis:

  • Manufacturer: Dell
  • Model: PowerEdge R630
  • Form Factor: 1U Rack Server
  • Rack Mountable: Yes

Processor:

  • CPU Model: [To be determined - typically Intel Xeon E5-2600 v3/v4 series]
  • CPU Count: 2 processors (dual socket)
  • CPU Cores: 10-18 cores per processor (20-36 total cores per system)
  • CPU Architecture: x86_64
  • CPU Speed: [To be determined]
  • Total CPU Cores (12 systems): 240-432 cores

Memory:

  • Total RAM per System: 128 GB
  • Total RAM (12 systems): 1,536 GB
  • RAM Type: DDR4 ECC RDIMM
  • Memory Slots: 24 DIMM slots (12 per CPU)
  • Memory Configuration: [To be determined]
  • Maximum Memory: Up to 1.5 TB (with LRDIMMs)

Storage:

  • Storage Controller: [To be determined - typically PERC H730/H730P]
  • Storage Bays: 10x 2.5" hot-swappable drive bays
  • Current Storage: [To be determined]
  • RAID Configuration: [To be determined]
  • Storage Options: SATA, SAS, NVMe (with riser card)

Network:

  • Network Adapters: [To be determined - typically 2x 1GbE onboard]
  • Network Ports: 2x 1GbE (onboard)
  • Additional NICs: [To be determined - supports PCIe NICs]
  • Network Options: 10GbE, 25GbE via PCIe cards

Power:

  • Power Supply: Dual redundant power supplies
  • Power Rating: 495W, 750W, or 1100W options
  • Power Consumption: [To be determined]

Proxmox VE Compatibility:

  • ✅ Fully Compatible - Dell R630 is fully supported by Proxmox VE
  • Recommended Proxmox VE Version: 8.x (latest stable)
  • Virtualization Support: Intel VT-x, VT-d
  • Storage: Supports local storage, Ceph, ZFS
  • Network: Supports standard network bridges, SR-IOV (with compatible NICs)

Deployment Notes:

  • Standard configuration suitable for general-purpose workloads
  • Excellent for Proxmox VE cluster nodes
  • Can be used for compute-intensive workloads
  • Ideal for distributed workloads across cluster
  • Consider for Ceph storage nodes (with additional storage)
  • Can be used for Kubernetes worker nodes

Cluster Recommendations:

  • These 12 systems are ideal for forming a Proxmox VE cluster
  • Recommended cluster size: 3-5 nodes for quorum
  • Can form multiple clusters or one large cluster
  • Consider Ceph storage cluster across these nodes

15. Dell Precision 7920 (High Memory + GPU)

System ID: PVE-HOST-015
Status: Active
Proxmox VE Version: [To be determined]
Cluster: [To be determined]

Hardware Specifications

Chassis:

  • Manufacturer: Dell
  • Model: Precision 7920 Tower
  • Form Factor: Tower Workstation/Server
  • Rack Mountable: Optional (with rack kit)

Processor:

  • CPU Model: [To be determined - typically Intel Xeon Scalable processors]
  • CPU Count: 2 processors (dual socket)
  • CPU Cores: 8-28 cores per processor (16-56 total cores)
  • CPU Architecture: x86_64
  • CPU Speed: [To be determined]

Memory:

  • Total RAM: 128 GB
  • RAM Type: DDR4 ECC
  • Memory Slots: [To be determined]
  • Memory Configuration: [To be determined]
  • Maximum Memory: Up to 3 TB (depending on configuration)

Graphics Processing Unit (GPU):

  • GPU Model: NVIDIA Quadro P5000
  • GPU Memory: 16 GB GDDR5X
  • GPU CUDA Cores: 2,560
  • GPU Architecture: Pascal (GP104)
  • GPU PCIe Slot: PCIe 3.0 x16
  • GPU Power: 180W TDP
  • GPU Features:
    • CUDA Compute Capability: 6.1
    • Supports GPU passthrough in Proxmox VE
    • Supports vGPU (with NVIDIA vGPU software)
    • Supports NVIDIA GRID virtualization

Storage:

  • Storage Controller: [To be determined]
  • Storage Bays: [To be determined]
  • Current Storage: [To be determined]
  • RAID Configuration: [To be determined]

Network:

  • Network Adapters: [To be determined]
  • Network Ports: [To be determined]
  • Additional NICs: [To be determined]

Power:

  • Power Supply: [To be determined]
  • Power Rating: [To be determined]
  • Power Consumption: [To be determined]

Proxmox VE Compatibility:

  • ✅ Fully Compatible - Dell Precision 7920 is fully supported by Proxmox VE
  • Recommended Proxmox VE Version: 8.x (latest stable)
  • Virtualization Support: Intel VT-x, VT-d (required for GPU passthrough)
  • GPU Passthrough: ✅ Supported (requires VT-d/IOMMU)
  • Storage: Supports local storage, Ceph, ZFS
  • Network: Supports standard network bridges

GPU Passthrough Configuration:

  • Requires IOMMU/VT-d enabled in BIOS
  • Requires proper PCIe passthrough configuration
  • Supports single GPU passthrough to one VM
  • Can use NVIDIA vGPU for multiple VMs (requires NVIDIA vGPU license)

Deployment Notes:

  • Ideal for GPU-accelerated workloads (AI/ML, rendering, compute)
  • Suitable for virtualized GPU workloads
  • Can host VMs requiring GPU acceleration
  • Consider for specialized workloads (rendering farms, AI training)
  • Excellent for development/testing GPU applications

16. Dell Precision 7920 (Standard Memory + GPU)

System ID: PVE-HOST-016
Status: Active
Proxmox VE Version: [To be determined]
Cluster: [To be determined]

Hardware Specifications

Chassis:

  • Manufacturer: Dell
  • Model: Precision 7920 Tower
  • Form Factor: Tower Workstation/Server
  • Rack Mountable: Optional (with rack kit)

Processor:

  • CPU Model: [To be determined - typically Intel Xeon Scalable processors]
  • CPU Count: 2 processors (dual socket)
  • CPU Cores: 8-28 cores per processor (16-56 total cores)
  • CPU Architecture: x86_64
  • CPU Speed: [To be determined]

Memory:

  • Total RAM: 64 GB
  • RAM Type: DDR4 ECC
  • Memory Slots: [To be determined]
  • Memory Configuration: [To be determined]
  • Maximum Memory: Up to 3 TB (depending on configuration)

Graphics Processing Unit (GPU):

  • GPU Model: NVIDIA Quadro P5000
  • GPU Memory: 16 GB GDDR5X
  • GPU CUDA Cores: 2,560
  • GPU Architecture: Pascal (GP104)
  • GPU PCIe Slot: PCIe 3.0 x16
  • GPU Power: 180W TDP
  • GPU Features:
    • CUDA Compute Capability: 6.1
    • Supports GPU passthrough in Proxmox VE
    • Supports vGPU (with NVIDIA vGPU software)
    • Supports NVIDIA GRID virtualization

Storage:

  • Storage Controller: [To be determined]
  • Storage Bays: [To be determined]
  • Current Storage: [To be determined]
  • RAID Configuration: [To be determined]

Network:

  • Network Adapters: [To be determined]
  • Network Ports: [To be determined]
  • Additional NICs: [To be determined]

Power:

  • Power Supply: [To be determined]
  • Power Rating: [To be determined]
  • Power Consumption: [To be determined]

Proxmox VE Compatibility:

  • ✅ Fully Compatible - Dell Precision 7920 is fully supported by Proxmox VE
  • Recommended Proxmox VE Version: 8.x (latest stable)
  • Virtualization Support: Intel VT-x, VT-d (required for GPU passthrough)
  • GPU Passthrough: ✅ Supported (requires VT-d/IOMMU)
  • Storage: Supports local storage, Ceph, ZFS
  • Network: Supports standard network bridges

GPU Passthrough Configuration:

  • Requires IOMMU/VT-d enabled in BIOS
  • Requires proper PCIe passthrough configuration
  • Supports single GPU passthrough to one VM
  • Can use NVIDIA vGPU for multiple VMs (requires NVIDIA vGPU license)

Deployment Notes:

  • Ideal for GPU-accelerated workloads (AI/ML, rendering, compute)
  • Suitable for virtualized GPU workloads
  • Can host VMs requiring GPU acceleration
  • Consider for specialized workloads (rendering farms, AI training)
  • Excellent for development/testing GPU applications
  • Lower memory configuration suitable for lighter GPU workloads

Available Hardware

Current Status

All 16 systems are currently in use. No additional hardware is available at this time.

Future Hardware Recommendations

For Cluster Expansion

  • Additional Dell R630 systems: For expanding compute cluster
  • Storage-optimized systems: For Ceph storage cluster expansion
  • Network switches: For improved network connectivity and redundancy

For GPU Workload Expansion

  • Additional GPU systems: For expanding GPU compute capacity
  • NVIDIA A100/H100 systems: For advanced AI/ML workloads
  • GPU servers: Dedicated GPU server systems

For Network Infrastructure

  • 10GbE/25GbE switches: For improved inter-node connectivity
  • Network adapters: 10GbE/25GbE PCIe cards for existing systems
  • Redundant network infrastructure: For high availability

Proxmox VE Cluster Configuration

Recommended Cluster Topology

Option 1: Single Large Cluster

  • Cluster Name: sankofa-pve-cluster-01
  • Nodes: All 16 systems
  • Quorum: 3-5 nodes for quorum (recommended: 5)
  • Storage: Ceph distributed storage across nodes
  • Network: Shared network infrastructure

Advantages:

  • Single management interface
  • Easy VM migration across all nodes
  • Centralized storage management
  • Simplified backup and disaster recovery

Considerations:

  • Requires reliable network connectivity
  • Quorum management with 16 nodes
  • Network bandwidth requirements

Option 2: Multiple Specialized Clusters

Cluster 1: Compute Cluster

  • Nodes: 1x HPE ML110 Gen9, 1x Dell R630 (768GB), 12x Dell R630 (128GB)
  • Total Nodes: 14
  • Purpose: General compute workloads
  • Storage: Ceph distributed storage

Cluster 2: GPU Cluster

  • Nodes: 2x Dell Precision 7920 (with NVIDIA P5000)
  • Total Nodes: 2
  • Purpose: GPU-accelerated workloads
  • Storage: Shared storage or local storage

Advantages:

  • Specialized clusters for different workloads
  • GPU cluster isolated for specialized workloads
  • Easier management of GPU resources

Considerations:

  • Multiple clusters to manage
  • Storage sharing between clusters
  • Network segmentation

Option 3: Hybrid Approach

Primary Cluster: 13x Dell R630 systems (compute cluster)

  • Nodes: 13 systems
  • Purpose: Primary compute infrastructure
  • Storage: Ceph distributed storage

Secondary Cluster: 1x HPE ML110 Gen9 + 2x Dell Precision 7920

  • Nodes: 3 systems
  • Purpose: Development/testing and GPU workloads
  • Storage: Local storage or shared storage

Advantages:

  • Separation of production and development
  • GPU resources in separate cluster
  • Flexible resource allocation

Storage Configuration

Recommended Storage Setup

Ceph Distributed Storage

  • Recommended Nodes: 6-12 Dell R630 systems
  • Storage Disks: [To be determined - requires inventory]
  • Network: Dedicated storage network (10GbE recommended)
  • Replication: 3x replication (recommended)
  • Pool Configuration:
    • RBD pool for VM disks
    • CephFS for shared filesystems
    • RGW for object storage (optional)

Local Storage

  • Use Cases:
    • OS disks for VMs
    • High-performance local storage
    • Backup storage
  • Recommended: ZFS on local storage for snapshots and compression

Shared Storage

  • Use Cases:
    • VM templates
    • ISO images
    • Backup storage
  • Options: NFS, CIFS, or CephFS

Network Configuration

Network Requirements

Management Network

  • Purpose: Proxmox VE management, cluster communication
  • Bandwidth: 1GbE minimum, 10GbE recommended
  • Redundancy: Bonded interfaces recommended

VM Network

  • Purpose: VM traffic, external connectivity
  • Bandwidth: 1GbE minimum, 10GbE recommended
  • VLANs: Recommended for network segmentation

Storage Network (Ceph)

  • Purpose: Ceph cluster communication, data replication
  • Bandwidth: 10GbE minimum, 25GbE recommended
  • Redundancy: Bonded interfaces required
  • Isolation: Dedicated network recommended

Migration Network

  • Purpose: Live migration traffic
  • Bandwidth: 10GbE recommended
  • Can share: With storage network or management network

Network Hardware Recommendations

Switches

  • Management/VM Network: 1GbE or 10GbE switches
  • Storage Network: 10GbE or 25GbE switches (dedicated)
  • Redundancy: Redundant switches for high availability

Network Adapters

  • Onboard NICs: Use for management network
  • PCIe NICs: 10GbE/25GbE cards for storage and VM networks
  • Bonding: Configure LACP bonds for redundancy

Proxmox VE Compatibility Matrix

Hardware Component Proxmox VE Support Notes
HPE ML110 Gen9 ✅ Full Support Standard x86_64 server
Dell R630 ✅ Full Support Enterprise server, excellent support
Dell Precision 7920 ✅ Full Support Workstation/server hybrid
Intel Xeon Processors ✅ Full Support All modern Xeon processors supported
DDR4 ECC Memory ✅ Full Support Recommended for Proxmox VE
NVIDIA P5000 GPU ✅ Full Support GPU passthrough supported
PERC RAID Controllers ✅ Full Support Use in HBA mode for ZFS/Ceph
Standard Network Adapters ✅ Full Support Intel, Broadcom, etc.

BIOS/UEFI Configuration Requirements

Required Settings for All Systems

Virtualization

  • ✅ Intel VT-x: Enable
  • ✅ Intel VT-d / IOMMU: Enable (required for GPU passthrough, PCIe passthrough)
  • ✅ SR-IOV: Enable (if supported and using SR-IOV)

CPU Settings

  • ✅ Hyperthreading: Enable (recommended)
  • ✅ CPU Power Management: Performance mode (recommended for servers)

Memory Settings

  • ✅ Memory ECC: Enable (if available)
  • ✅ NUMA: Enable (for multi-socket systems)

Storage Settings

  • ✅ AHCI Mode: For ZFS/Ceph (if not using hardware RAID)
  • ✅ RAID Mode: For hardware RAID (if using hardware RAID)

Boot Settings

  • ✅ UEFI Boot: Enable (recommended)
  • ✅ Secure Boot: Disable (for Proxmox VE compatibility)
  • ✅ Legacy Boot: Disable (if using UEFI)

Performance Recommendations

CPU Allocation

  • Host CPU Reservation: Reserve 1-2 cores per host for Proxmox VE
  • VM CPU Allocation: Use CPU pinning for performance-critical VMs
  • NUMA Awareness: Configure NUMA for multi-socket systems

Memory Allocation

  • Host Memory Reservation: Reserve 4-8 GB per host for Proxmox VE
  • Balloon Driver: Enable for memory overcommitment
  • Memory Hotplug: Enable for dynamic memory allocation

Storage Performance

  • Use NVMe/SSD: For VM disks requiring high IOPS
  • Use Ceph: For distributed storage and high availability
  • Use ZFS: For local storage with snapshots and compression
  • RAID Configuration: RAID 10 for performance, RAID 5/6 for capacity

Network Performance

  • Use 10GbE/25GbE: For storage and migration networks
  • Enable Jumbo Frames: For storage network (MTU 9000)
  • Use SR-IOV: For high-performance network requirements

Monitoring and Management

Recommended Tools

  • Proxmox VE Web Interface: Primary management interface
  • Proxmox VE CLI: Command-line management
  • Prometheus + Grafana: Monitoring and alerting
  • Zabbix: Alternative monitoring solution
  • Proxmox Backup Server: Backup and disaster recovery

Key Metrics to Monitor

  • CPU Usage: Per host and per VM
  • Memory Usage: Per host and per VM
  • Storage Usage: Per storage pool and per VM
  • Network Usage: Per interface and per VM
  • Cluster Health: Quorum status, node status
  • Ceph Health: Cluster status, OSD status, pool usage

Backup and Disaster Recovery

Backup Strategy

  • Proxmox Backup Server: Recommended for centralized backups
  • VM Backups: Full backups and incremental backups
  • Backup Frequency: Daily backups recommended
  • Retention Policy: 30-90 days recommended

Disaster Recovery

  • Cluster Configuration: Backup cluster configuration
  • VM Templates: Backup VM templates and ISOs
  • Storage Configuration: Document storage setup
  • Network Configuration: Document network setup
  • Recovery Procedures: Document recovery procedures

Security Considerations

Proxmox VE Security

  • Firewall: Enable Proxmox VE firewall
  • SSH Access: Restrict SSH access, use key-based authentication
  • Web Interface: Use HTTPS, restrict access
  • API Access: Use API tokens, restrict permissions
  • Updates: Regular security updates

VM Security

  • Guest Agent: Install QEMU guest agent in VMs
  • Firewall: Configure firewall in VMs
  • Updates: Regular security updates in VMs
  • Access Control: Use Proxmox VE user management

Next Steps

Immediate Actions

  1. Hardware Inventory: Complete detailed hardware inventory (CPU models, storage, network)
  2. BIOS Configuration: Configure BIOS/UEFI settings on all systems
  3. Proxmox VE Installation: Install Proxmox VE on all systems
  4. Cluster Formation: Form Proxmox VE cluster(s)
  5. Network Configuration: Configure network interfaces and bonds
  6. Storage Configuration: Configure storage (Ceph, local, shared)
  7. Testing: Test cluster functionality, VM creation, migration

Future Enhancements

  1. Storage Expansion: Add additional storage to Ceph cluster
  2. Network Upgrades: Upgrade to 10GbE/25GbE for storage network
  3. GPU Passthrough: Configure GPU passthrough on Precision 7920 systems
  4. Monitoring Setup: Deploy monitoring and alerting
  5. Backup Setup: Deploy Proxmox Backup Server
  6. Documentation: Complete detailed documentation

Related Documentation

Last Updated: [Date]
Status: In Progress
Maintainer: Infrastructure Team
Version: 1.0

Reference in New Issue View Git Blame Copy Permalink