Avidys · Servers · Docs · SvelteTV · ClinStats
Server benchmarks
Last updated 2026-11-07 03:14 UTC · refreshes every 15 minutes (page auto-reloads)
CPU bench is SHA-256 throughput over 5 s (single-core, then all cores). Disk bench is a 1 GB
file: sequential write with fsync, sequential read with direct I/O. Network bench is a 50 MB download
from Cloudflare. Usage (load, memory, network rates, top processes) is a live snapshot from collection time.
The Services section in each machine's card lists what is installed and running on it.
At a glance
Tap any column header for what it measures. Docs — methods & publishing →
Capacity — what the machine is
Current — what it is doing now
PassMark CPU Mark is an independent benchmark from cpubenchmark.net, checked
2026-11-06, per CPU model (per socket). Dual-socket machines (×2) score higher as a whole
system — e.g. a dual Xeon E5-2690 v4 system scores 30,555. Main disk is the whole physical
disk holding the root filesystem; its used figure sums every local filesystem on that disk.
What do these numbers mean?
Capacity — the machine's fixed specs: CPU model and frequency, cores and
threads, total RAM, and the whole capacity of its main disk. These don't change between runs.
Current — live measurements from the last collection: CPU load, RAM and disk
use as percentages, plus the latest benchmark runs (CPU multi, disk write/read, network download).
Main disk — the physical disk holding the root filesystem. The root
filesystem is only a partition of it, so the disk's total is bigger than the / size; used is
the sum of every local filesystem on that disk.
PassMark — an independent CPU benchmark from cpubenchmark.net for the CPU
model (checked 2026-11-06). It lets you compare these machines with any other CPU on the
same scale; the SHA-256 multi-core figure next to it is our own quick measurement.
Services — what is installed and running on top of the base system:
the running services with a one-line description, any failed units (marked ⚠️), and
Docker containers when Docker is installed.
CPU load % — the 1-minute load average divided by the thread count.
100% means every thread is busy; compare machines of different sizes fairly.
Disk write / read (MB/s) — sequential transfer of a 1 GB test file. The read
uses direct I/O so it measures the disk, not the memory cache. The test file is deleted afterwards.
Net down (Mbps) — a 50 MB download from Cloudflare; depends on the server's
internet link at that moment.
Full methods and the publishing pipeline: docs .
.6 debian-6 — via LAN subnet route (debian8)
Debian GNU/Linux 12 (bookworm) · kernel 6.1.0-22-amd64 · uptime 2.3 days
· collected 2026-11-07 03:14 UTC
LAN/tailnet: 192.168.68.6, 172.17.0.1 · interfaces: enp1s0 (192.168.68.6/24)
Hardware & capacity
Processor Intel(R) N100
Cores / threads 4 cores · 4 threads · 1 socket(s) · 3,400 MHz max · L3 6 MiB (1 instance)
CPU bench (SHA-256) single-core 10,465 MB/s · multi-core (4 workers) 35,274 MB/s
Memory 15 GB total · write 5,582 MB/s · copy 4,711 MB/s
Main disk (holds /) /dev/nvme0n1 — 477 GB total · 27 GB used (5.7%)
Disk bench (1 GB file) write 511 MB/s · read (direct I/O) 685 MB/s
Network bench download 103 Mbps · latency 68.0 ms
Disks
Device Size Model Transport /dev/nvme0n1 477 GB 512GB SSD nvme
Filesystems
Mount Source Size Used / /dev/nvme0n1p2 467 GB 7% /boot/efi /dev/nvme0n1p1 0 GB 2%
Current usage
Load (1 / 5 / 15 min) 1.24 / 0.35 / 0.22
Memory 1 GB used of 15 GB (14 GB available) 7%
Services
Service What it does containerd.service containerd container runtime docker.service Docker Application Container Engine named.service BIND Domain Name Server netdata.service Netdata, X-Ray Vision for your infrastructure! openclaw-gateway.service OpenClaw Gateway (v2026.4.12)
Docker is running, no containers.
Network right now (1 s sample)
Interface Down Up Total transferred enp1s0 ↓ 13.8 KB/s ↑ 40.0 KB/s ↓ 3.31 GB · ↑ 4.2 GB
Top processes
Process CPU Memory python3 27.5% 0.1% netdata 5.4% 0.9% apps.plugin 0.9% 0.0% go.d.plugin 0.4% 0.7% openclaw-gatewa 0.3% 2.1%
.7 debian-7 — via LAN subnet route (debian8)
Debian GNU/Linux 12 (bookworm) · kernel 6.1.0-22-amd64 · uptime 9.1 days
· collected 2026-11-07 03:14 UTC
LAN/tailnet: 192.168.68.7 · interfaces: eno1 (192.168.68.7/24)
Hardware & capacity
Processor AMD Ryzen 5 5500U with Radeon Graphics
Cores / threads 6 cores · 12 threads · 1 socket(s) · 4,056 MHz max · L3 8 MiB (2 instances)
CPU bench (SHA-256) single-core 9,513 MB/s · multi-core (12 workers) 85,006 MB/s
Memory 15 GB total · write 7,218 MB/s · copy 3,339 MB/s
Main disk (holds /) /dev/nvme0n1 — 477 GB total · 22 GB used (4.5%)
Disk bench (1 GB file) write 757 MB/s · read (direct I/O) 2,200 MB/s
Network bench download 44 Mbps · latency 37.7 ms
Disks
Device Size Model Transport /dev/sda 477 GB SATA SSD sata /dev/nvme0n1 477 GB Lexar SSD NM6A1 512GB nvme
Filesystems
Mount Source Size Used / /dev/nvme0n1p2 467 GB 5% /boot/efi /dev/nvme0n1p1 0 GB 2%
Current usage
Load (1 / 5 / 15 min) 1.51 / 0.63 / 0.5
Memory 10 GB used of 15 GB (5 GB available) 68%
Services
Service What it does bbs-agent.service Borg Backup Server Agent gitlab-runsvdir.service GitLab Runit supervision process netdata.service Netdata, X-Ray Vision for your infrastructure!
Network right now (1 s sample)
Interface Down Up Total transferred eno1 ↓ 13.9 KB/s ↑ 40.0 KB/s ↓ 3.45 GB · ↑ 7.7 GB
Top processes
Process CPU Memory python3 23.6% 0.1% netdata 9.5% 1.3% ruby 7.8% 7.6% apps.plugin 2.3% 0.0% ruby 1.8% 0.5%
.8 debian8 — direct SSH from this VM
Debian GNU/Linux 12 (bookworm) · kernel 6.1.0-28-amd64 · uptime 9.1 days
· collected 2026-11-07 03:12 UTC
LAN/tailnet: 192.168.68.8, 100.78.57.116, fd7a:115c:a1e0::7001:3977 · interfaces: eno1 (192.168.68.8/24)
Hardware & capacity
Processor AMD Ryzen 7 5825U with Radeon Graphics
Cores / threads 8 cores · 16 threads · 1 socket(s) · 4,547 MHz max · L3 16 MiB (1 instance)
CPU bench (SHA-256) single-core 10,483 MB/s · multi-core (16 workers) 115,792 MB/s
Memory 15 GB total · write 6,923 MB/s · copy 4,805 MB/s
Main disk (holds /) /dev/sda — 477 GB total · 72 GB used (15.1%)
Disk bench (1 GB file) write 313 MB/s · read (direct I/O) 526 MB/s
Network bench download 103 Mbps · latency 27.8 ms
Disks
Device Size Model Transport /dev/sda 477 GB UDSS UD2CS1HT100-512G sata /dev/sdb 3,686 GB Expansion Desk usb
Filesystems
Mount Source Size Used / /dev/sda2 23 GB 70% /tmp /dev/sda5 2 GB 12% /var /dev/sda3 9 GB 28% /boot/efi /dev/sda1 0 GB 2% /home /dev/sda6 433 GB 14% /mnt/Bots //192.168.68.5/Bots 3,663 GB 87% /mnt/NAS/Avidys //192.168.68.5/Avidys 3,663 GB 87%
Current usage
Load (1 / 5 / 15 min) 1.77 / 0.66 / 0.39
Memory 4 GB used of 15 GB (11 GB available) 26%
Services
Service What it does bbs-agent.service Borg Backup Server Agent dns.service Technitium DNS Server glances.service Glances lighttpd.service Lighttpd Daemon lsws.service LSB: lshttpd netdata.service Netdata, X-Ray Vision for your infrastructure! ollama.service Ollama Service plexmediaserver.service Plex Media Server snmpd.service Simple Network Management Protocol (SNMP) Daemon. [email protected] Syncthing - Open Source Continuous File Synchronization for avidys tailscaled.service Tailscale node agent openclaw-gateway.service OpenClaw Gateway (v2026.4.15)
⚠️ Failed: logrotate.service — Rotate log files; pomerium.service — Pomerium
Network right now (1 s sample)
Interface Down Up Total transferred eno1 ↓ 91.0 KB/s ↑ 38.3 KB/s ↓ 16.44 GB · ↑ 4.82 GB tailscale0 ↓ 0 B/s ↑ 0 B/s ↓ 0.02 GB · ↑ 0.02 GB
Top processes
Process CPU Memory python3 30.9% 0.1% netdata 4.8% 2.6% apps.plugin 2.9% 0.0% go.d.plugin 1.8% 1.6%
.15 debian-15 — direct SSH from this VM
Debian GNU/Linux 13 (trixie) · kernel 6.12.107+deb13-amd64 · uptime 2.3 days
· collected 2026-11-07 03:12 UTC
LAN/tailnet: 192.168.68.15, 100.89.130.42, fd7a:115c:a1e0::5032:822b · interfaces: enp2s0 (192.168.68.15/24)
Hardware & capacity
Processor Intel(R) N100
Cores / threads 4 cores · 4 threads · 1 socket(s) · 3,400 MHz max · L3 6 MiB (1 instance)
CPU bench (SHA-256) single-core 10,508 MB/s · multi-core (4 workers) 35,438 MB/s
Memory 15 GB total · write 5,254 MB/s · copy 3,954 MB/s
Main disk (holds /) /dev/nvme0n1 — 954 GB total · 30 GB used (3.2%)
Disk bench (1 GB file) write 2,800 MB/s · read (direct I/O) 9,000 MB/s
Network bench download 84 Mbps · latency 116.1 ms
Disks
Device Size Model Transport /dev/nvme0n1 954 GB RS1D0TSSD710 nvme
Filesystems
Mount Source Size Used / /dev/mapper/debian--15--vg-root 53 GB 58% /sys/firmware/efi/efivars efivarfs 0 GB 53% /boot /dev/nvme0n1p2 1 GB 31% /boot/efi /dev/nvme0n1p1 1 GB 1% /var /dev/mapper/debian--15--vg-var 21 GB 6% /srv /dev/mapper/debian--15--vg-srv 840 GB 1%
Current usage
Load (1 / 5 / 15 min) 0.39 / 0.2 / 0.18
Memory 1 GB used of 15 GB (14 GB available) 5%
Services
Service What it does caddy.service Caddy cloudflared.service Cloudflare Tunnel client ollama.service Ollama Service tailscaled.service Tailscale node agent hermes-gateway-coder.service Hermes Agent Gateway - Messaging Platform Integration hermes-gateway.service Hermes Agent Gateway - Messaging Platform Integration nanobot-gateway.service Nanobot Gateway (nanobot-gateway) syncthing.service Syncthing - Open Source Continuous File Synchronization zeroclaw.service ZeroClaw daemon
Network right now (1 s sample)
Interface Down Up Total transferred tailscale0 ↓ 0 B/s ↑ 0 B/s ↓ 0.0 GB · ↑ 0.01 GB enp2s0 ↓ 626 B/s ↑ 534 B/s ↓ 1.93 GB · ↑ 0.44 GB
Top processes
Process CPU Memory python3 31.2% 0.1% python 1.5% 0.7% hermes 0.5% 0.8% cloudflared 0.4% 0.3% hermes 0.4% 0.8% tailscaled 0.3% 0.3%
.21 dell-r630-2x — via LAN subnet route (debian8)
Ubuntu 24.04.4 LTS · kernel 6.8.0-137-generic · uptime 87.2 days
· collected 2026-11-07 03:13 UTC
LAN/tailnet: 192.168.68.21, 192.168.68.22 · interfaces: eno1 (192.168.68.21/24), eno2 (192.168.68.22/24)
Hardware & capacity
Processor Intel(R) Xeon(R) CPU E5-2690 v4 @ 2.60GHz
Cores / threads 28 cores · 56 threads · 2 socket(s) · 3,500 MHz max · L3 70 MiB (2 instances)
CPU bench (SHA-256) single-core 1,868 MB/s · multi-core (56 workers) 54,188 MB/s
Memory 126 GB total · write 1,755 MB/s · copy 1,528 MB/s
Main disk (holds /) /dev/sdb — 953 GB total · 117 GB used (12.3%)
Disk bench (1 GB file) write 663 MB/s · read (direct I/O) 894 MB/s
Network bench download 44 Mbps · latency 37.6 ms
Disks
Device Size Model Transport /dev/sda 112 GB Samsung SSD 840 — /dev/sdb 953 GB PERC H730 Mini — /dev/sdc 9,318 GB easystore 25FB usb
Filesystems
Mount Source Size Used / /dev/sdb2 937 GB 14% /mnt/WD9T /dev/sdc1 9,314 GB 26% /mnt/WD4G/OLD //192.168.68.5/OLD 3,663 GB 87% /mnt/WD4G/public //192.168.68.5/public 3,663 GB 87%
Current usage
Load (1 / 5 / 15 min) 3.62 / 0.96 / 0.46
Memory 5 GB used of 126 GB (123 GB available) 4%
Services
Service What it does atop.service Atop advanced performance monitor atopacct.service Atop process accounting daemon cloudflared.service cloudflared glances.service Glances ollama.service Ollama Service rstudio-server.service RStudio Server snap.nextcloud.apache.service Service for snap application nextcloud.apache snap.nextcloud.mysql.service Service for snap application nextcloud.mysql snap.nextcloud.nextcloud-cron.service Service for snap application nextcloud.nextcloud-cron snap.nextcloud.php-fpm.service Service for snap application nextcloud.php-fpm snap.nextcloud.redis-server.service Service for snap application nextcloud.redis-server snap.nextcloud.renew-certs.service Service for snap application nextcloud.renew-certs
⚠️ Failed: isc-dhcp-server6.service — ISC DHCP IPv6 server; logrotate.service — Rotate log files; snap.ollama.listener.service — Service for snap application ollama.listener; systemd-networkd-wait-online.service — Wait for Network to be Configured
Network right now (1 s sample)
Interface Down Up Total transferred eno2 ↓ 120 B/s ↑ 0 B/s ↓ 2.82 GB · ↑ 0.01 GB eno1 ↓ 646 B/s ↑ 968 B/s ↓ 35.36 GB · ↑ 6.42 GB
Top processes
Process CPU Memory python3 23.7% 0.0% systemd 0.9% 0.0% mysqld 0.8% 0.2% redis-server 0.5% 0.0% cloudflared 0.2% 0.0%
.43 r630-4x — direct SSH from this VM
Debian GNU/Linux 13 (trixie) · kernel 6.12.73+deb13-amd64 · uptime 87.2 days
· collected 2026-11-07 03:13 UTC
LAN/tailnet: 192.168.68.43, 192.168.68.44, 100.85.1.101, fd7a:115c:a1e0::b332:165 · interfaces: eno3 (192.168.68.43/24), eno4 (192.168.68.44/24)
Hardware & capacity
Processor Intel(R) Xeon(R) CPU E5-2640 v3 @ 2.60GHz
Cores / threads 16 cores · 32 threads · 2 socket(s) · 3,400 MHz max · L3 40 MiB (2 instances)
CPU bench (SHA-256) single-core 1,576 MB/s · multi-core (32 workers) 26,964 MB/s
Memory 63 GB total · write 3,206 MB/s · copy 2,859 MB/s
Main disk (holds /) /dev/sda — 476 GB total · 107 GB used (22.5%)
Disk bench (1 GB file) write 2,200 MB/s · read (direct I/O) 6,600 MB/s
Network bench download 38 Mbps · latency 167.3 ms
Disks
Device Size Model Transport /dev/sda 476 GB PERC H330 Mini —
Filesystems
Mount Source Size Used / /dev/sda1 467 GB 25% /mnt/WD4G/public //192.168.68.5/public 3,663 GB 87% /mnt/WD4G/OLD //192.168.68.5/OLD 3,663 GB 87%
Current usage
Load (1 / 5 / 15 min) 3.27 / 1.29 / 0.94
Memory 4 GB used of 63 GB (59 GB available) 6%
Services
Service What it does exim4.service exim Mail Transport Agent mariadb.service MariaDB 10.11.6 database server nanobot.service Nanobot Gateway nginx.service A high performance web server and a reverse proxy server ollama.service Ollama Service php8.2-fpm.service The PHP 8.2 FastCGI Process Manager tailscaled.service Tailscale node agent
⚠️ Failed: certbot.service — Certbot; gitlab-runsvdir.service — GitLab Runit supervision process; [email protected] — ifup for eno3
Network right now (1 s sample)
Interface Down Up Total transferred eno3 ↓ 510 B/s ↑ 0 B/s ↓ 3.51 GB · ↑ 0.0 GB eno4 ↓ 343 B/s ↑ 799 B/s ↓ 33.22 GB · ↑ 3.36 GB
Top processes
Process CPU Memory python3 23.5% 0.0% tailscaled 0.1% 0.1% sshd-session 0.1% 0.0% nanobot 0.1% 0.3% rcu_preempt 0.0% 0.0%
Not benchmarked
mac-mini-3 — online on the tailnet, but SSH is closed (Remote Login likely off).4 — 192.168.68.4 on the LAN, but SSH is closed/unreachablem3pro-laptop — offline — last seen 51 days agom1pro-laptop — offline — last seen 111 days agoiphone-13-mini — offline — last seen 411 days ago
Benchmarks are collected by direct SSH from this VM, as the avidys user
— over the tailnet for debian-15, debian8 and r630-4x, and through debian8's subnet route
for the LAN machines. Collected with bench.py (no root, no extra packages);
the 1 GB disk-test file is deleted after each run.