feat: add VM sizing, host hardware specs and Proxmox provisioning steps

2026-05-29 08:00:52 +02:00
parent 4295bec6c0
commit 68f69c30ea
1 changed files with 174 additions and 0 deletions
--- a/README.md
+++ b/README.md
@@ -150,6 +150,180 @@ Telegraf, Beszel

 ---

+## Host Hardware
+
+| Host | CPU | Logical CPUs | RAM | Storage | VM |
+|---|---|---|---|---|---|
+| dracula host | Xeon E5-1620 v2 @ 3.70GHz | 8 (4 cores + HT) | 64GB | Datacenter HDD | dracula-new (web) |
+| transilvan host | Xeon E5-1620 v2 @ 3.70GHz | 8 (4 cores + HT) | 32GB | SSD | transilvan-new (db) |
+
+Both hosts run Proxmox VE (confirmed via qemu-guest-agent on current VMs). New VMs are the sole tenants on their respective hosts.
+
+---
+
+## VM Sizing
+
+### Rationale for 200 concurrent users
+
+With nginx fastcgi_cache, the vast majority of requests never invoke PHP. On a news site, traffic concentrates on recent articles — the hot working set is small and stays warm in FreeBSD's unified buffer cache (UBC) regardless of underlying disk speed.
+
+| Metric | Value | Notes |
+|---|---|---|
+| Expected cache HIT rate | ~85–90% | News sites have concentrated traffic on recent content |
+| Concurrent PHP requests at peak | ~20–30 | 200 users × ~15% miss rate |
+| PHP-FPM workers (pm.max_children) | 40 | Headroom above expected peak |
+| RAM per PHP-FPM worker | ~35MB RSS | WordPress with active plugins |
+| PHP-FPM peak RAM | ~1.4GB | 40 workers × 35MB |
+| nginx HIT response time | <5ms | Served from UBC, no PHP involved |
+| nginx MISS response time | 100–400ms | WP query + render |
+| InnoDB buffer pool hit rate | >99% | 20GB pool, 9.6GB dataset fits entirely in RAM |
+| Active DB connections | ≤ 40 | One per active PHP-FPM worker |
+
+### dracula-new (web VM on HDD host)
+
+| Parameter | Value |
+|---|---|
+| vCPU | 6 |
+| RAM | 14GB |
+| Disk | 80GB single virtual disk (HDD-backed) |
+| Balloon | Enabled (web VM, acceptable) |
+
+RAM breakdown: PHP-FPM peak 1.4GB + Redis 512MB + nginx + UBC page cache for hot articles (~4–6GB effective) + OS 1GB + headroom.
+
+HDD is acceptable for the web node. The fastcgi_cache working set for a news site fits in UBC (RAM). HDD latency only affects cold cache startup and log writes (sequential). The CPU and PHP-FPM worker slots are the actual bottleneck, not disk.
+
+### transilvan-new (DB VM on SSD host)
+
+| Parameter | Value |
+|---|---|
+| vCPU | 4 |
+| RAM | 26GB |
+| Disk 1 — OS | 40GB virtual (SSD-backed) |
+| Disk 2 — ZFS data pool | 150GB virtual (SSD-backed) |
+| Balloon | **Disabled** — memory ballooning must not reclaim InnoDB buffer pool RAM |
+
+RAM breakdown: innodb_buffer_pool_size 20GB + MySQL overhead 2GB + OS 4GB.
+
+Both virtual disks originate from the same physical SSD pool in Proxmox. Separation is logical: independent sizing, clean `zpool create data /dev/da1`, ZFS snapshots on the data disk without touching the OS disk, and easy future migration if a dedicated physical disk is added to the host.
+
+---
+
+## VM Provisioning (Proxmox)
+
+### 1. Download FreeBSD 14.2 ISO on each Proxmox host
+
+```sh
+wget -P /var/lib/vz/template/iso/ \
+  https://download.freebsd.org/releases/amd64/amd64/ISO-IMAGES/14.2/FreeBSD-14.2-RELEASE-amd64-disc1.iso
+```
+
+### 2. Create dracula-new (run on the HDD host)
+
+Via CLI:
+```sh
+qm create 200 \
+  --name dracula-new \
+  --memory 14336 \
+  --balloon 14336 \
+  --cores 6 \
+  --cpu host \
+  --machine q35 \
+  --bios ovmf \
+  --efidisk0 local:1,format=raw \
+  --net0 virtio,bridge=vmbr0 \
+  --ostype other \
+  --scsihw virtio-scsi-single \
+  --scsi0 local:80,format=raw \
+  --cdrom local:iso/FreeBSD-14.2-RELEASE-amd64-disc1.iso \
+  --boot order=ide2;scsi0 \
+  --agent enabled=1
+```
+
+Via UI:
+```
+General   → Name: dracula-new
+OS        → FreeBSD (other), ISO: FreeBSD-14.2-RELEASE-amd64-disc1.iso
+System    → Machine: q35, BIOS: OVMF (UEFI), Qemu Agent: ✓
+Disks     → VirtIO SCSI, 80GB, Cache: None
+CPU       → 6 cores, Type: host
+Memory    → 14336 MB
+Network   → VirtIO, vmbr0
+```
+
+### 3. Create transilvan-new (run on the SSD host)
+
+```sh
+# VM with OS disk
+qm create 201 \
+  --name transilvan-new \
+  --memory 26624 \
+  --balloon 0 \
+  --cores 4 \
+  --cpu host \
+  --machine q35 \
+  --bios ovmf \
+  --efidisk0 local:1,format=raw \
+  --net0 virtio,bridge=vmbr0 \
+  --ostype other \
+  --scsihw virtio-scsi-single \
+  --scsi0 local:40,format=raw,discard=on,ssd=1 \
+  --cdrom local:iso/FreeBSD-14.2-RELEASE-amd64-disc1.iso \
+  --boot order=ide2;scsi0 \
+  --agent enabled=1
+
+# Add second disk for ZFS data pool
+qm set 201 --scsi1 local:150,format=raw,discard=on,ssd=1
+```
+
+`--balloon 0` disables memory ballooning on the DB VM. This is required — ballooning can silently reclaim pages from the InnoDB buffer pool under host memory pressure.
+
+### 4. FreeBSD installer settings (same for both VMs)
+
+```
+Welcome screen   → Install
+Keymap           → your preference
+Hostname         → dracula-new  (or transilvan-new)
+Distribution     → base, kernel  (nothing else)
+Partitioning     → Auto (ZFS)
+  Pool type      → stripe (single disk)
+  Disk           → da0  (the OS disk)
+  Swap           → 2GB
+  Compress       → lz4
+  Encrypt        → No
+Network          → vtnet0, configure with a temporary IP for initial pkg bootstrap
+                   (Tailscale will take over; this IP can be removed afterwards)
+Mirror           → closest region
+Root password    → set strong password
+SSHD             → enable
+```
+
+### 5. First boot — both VMs
+
+```sh
+# Update base system
+freebsd-update fetch install
+
+# Bootstrap pkg and install essentials
+pkg update && pkg upgrade -y
+pkg install -y qemu-guest-agent tailscale sudo curl wget
+
+# Enable QEMU guest agent (so Proxmox can see VM IP, issue graceful shutdowns)
+sysrc qemu_guest_agent_enable=YES
+service qemu-guest-agent start
+
+# Join Tailscale network
+sysrc tailscaled_enable=YES
+service tailscaled start
+tailscale up --hostname=dracula-new    # or transilvan-new
+
+# Note the assigned Tailscale IP — plug into pf.conf and wp-config placeholders
+tailscale ip -4
+```
+
+After both VMs have Tailscale IPs, replace all `<dracula-new Tailscale IP>` and `<transilvan-new Tailscale IP>` placeholders throughout this document and in all config files.
+
+---
+
 ## Migration Data Inventory

 | Component | Size | Action |