# Hetzner CPX02 [Singapore](https://rackcensus.com/locations/singapore-vps-benchmarks "Every provider in Singapore") Benchmark complete Finished Oct 22, 2026 Geekbench 7 single-core Skipped Geekbench 7 multi-core Skipped Disk 4K IOPS 72,145 fio random read and write, 50/50 Monthly price at test time $8.99 Same price today ## Server [Hetzner](https://rackcensus.com/vps-providers/hetzner-benchmarks)[CPX02](https://rackcensus.com/plans/c2ynad-hetzner-cpx02-benchmarks?location=sin) - Provider Hetzner - Plan CPX02 - Location Singapore - vCPU 1 - RAM 1 GB - Storage 20 GB NVMe - Network Unknown - OS image ubuntu-24.04 - Trigger Scheduled - Queued at Oct 21, 2026, 00:20:00 UTC [Visit Hetzner](https://rackcensus.com/go/hetzner) ## Latest runs of this plan - [Helsinki DC Park 1 Oct 19, 2026](https://rackcensus.com/benchmarks/pkld1z-hetzner-cpx02-hel1-benchmark)Skipped - [Falkenstein DC Park 1 Oct 19, 2026](https://rackcensus.com/benchmarks/s11u07-hetzner-cpx02-fsn1-benchmark)Skipped [Full history of this plan](https://rackcensus.com/plans/c2ynad-hetzner-cpx02-benchmarks/history) ## CPU Geekbench was skipped because this plan has less than 2 GB of RAM, which isn't enough for Geekbench 7 to run without swap. ## Disk | Block size | Read | Write | Total | | --- | --- | --- | --- | | 4K | 140.7 MB/s (36,025 IOPS) | 141.1 MB/s (36,120 IOPS) | 281.8 MB/s (72,145 IOPS) | | 64K | 1.26 GB/s (20,567 IOPS) | 1.26 GB/s (20,675 IOPS) | 2.52 GB/s (41,242 IOPS) | | 512K | 2.17 GB/s (4,438 IOPS) | 2.28 GB/s (4,674 IOPS) | 4.45 GB/s (9,112 IOPS) | | 1M | 2.29 GB/s (2,342 IOPS) | 2.44 GB/s (2,498 IOPS) | 4.73 GB/s (4,840 IOPS) | fio random read and write, 50/50, run by YABS against the boot disk. ## Network ### IPv4 transfer speed Loading... | Target | Protocol | Send | Receive | | --- | --- | --- | --- | | Leaseweb, Singapore, SG (10G) | IPv4 | 9.28 Gbps | 9.36 Gbps | | Eranium, Amsterdam, NL (100G) | IPv4 | 850 Mbps | 1.44 Gbps | | Leaseweb, NYC, NY, US (10G) | IPv4 | 768 Mbps | 968 Mbps | Public iperf3 servers vary in capacity and availability. Compare a target only with itself over time. ## Application workloads ### PostgreSQL transactions Failed Durable transfers and indexed account lookups against a fixed relational dataset. RuntimeError: TimeoutError: load generator worker did not respond p95 ≤ 50 ms · p99 ≤ 100 ms · errors ≤ 1.0% ### PostgreSQL large dataset Insufficient RAM The same transaction mix against ten million accounts; the dataset never shrinks to fit a plan. Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. p95 ≤ 100 ms · p99 ≤ 250 ms · errors ≤ 1.0% ### Redis cache Measured A mixed cache workload with 1 KiB values, expiration, misses and eviction. Redis cache measurements| Repeat | Offered / s | operations/s | p95 / p99 ms | Errors / dropped | Result | | --- | --- | --- | --- | --- | --- | | 1 | 1000 | 1000 | 1.84 / 2.05 | 0 / 0 | Met workload limits | | 1 | 5000 | 4999.99 | 1.93 / 2.27 | 0 / 0 | Met workload limits | | 1 | 10000 | 9998.8 | 1.31 / 1.93 | 0 / 212 | Met workload limits | | 1 | 25000 | 23372 | 19.19 / 25.67 | 0 / 292785 | load generator saturated | | 2 | 1000 | 999.99 | 1.9 / 2.05 | 0 / 0 | Met workload limits | | 2 | 5000 | 4999.97 | 1.96 / 2.3 | 0 / 0 | Met workload limits | | 2 | 10000 | 9999.07 | 1.28 / 1.93 | 0 / 163 | Met workload limits | | 2 | 25000 | 23354.4 | 19.39 / 25.41 | 0 / 295926 | load generator saturated | p95 ≤ 10 ms · p99 ≤ 25 ms · errors ≤ 1.0% ### Redis persistent queue Measured Enqueue, consume and acknowledge payloads with append-only persistence. Redis persistent queue measurements| Repeat | Offered / s | jobs/s | p95 / p99 ms | Errors / dropped | Result | | --- | --- | --- | --- | --- | --- | | 1 | 100 | 100 | 4.27 / 4.82 | 0 / 0 | Met workload limits | | 1 | 500 | 500 | 3.32 / 4.11 | 0 / 0 | Met workload limits | | 1 | 1000 | 999.99 | 3.36 / 4.32 | 0 / 0 | Met workload limits | | 1 | 3000 | 2999.82 | 3.36 / 5.76 | 0 / 29 | Met workload limits | | 2 | 100 | 100 | 4.32 / 4.94 | 0 / 0 | Met workload limits | | 2 | 500 | 500 | 3.23 / 4.01 | 0 / 0 | Met workload limits | | 2 | 1000 | 999.99 | 3.36 / 4.27 | 0 / 0 | Met workload limits | | 2 | 3000 | 2999.96 | 3.32 / 5.82 | 0 / 3 | Met workload limits | p95 ≤ 20 ms · p99 ≤ 50 ms · errors ≤ 1.0% ### Discourse community Insufficient RAM Signed-in reading, search, posting and replies in a populated Rails forum. Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. p95 ≤ 1500 ms · p99 ≤ 3000 ms · errors ≤ 1.0% ### n8n workflows Insufficient RAM Webhook, transformation, branching and a verified final response. Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. p95 ≤ 1000 ms · p99 ≤ 2000 ms · errors ≤ 1.0% ### n8n queue workers Insufficient RAM Completed queued workflows including waiting time and worker execution. Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. p95 ≤ 2000 ms · p99 ≤ 4000 ms · errors ≤ 1.0% ### Ghost publishing Insufficient RAM Readers, authenticated draft creation, publishing and image uploads. Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. p95 ≤ 1000 ms · p99 ≤ 2000 ms · errors ≤ 1.0% ### Formbricks / Next.js Insufficient RAM Public survey rendering, persisted responses and authenticated analysis. Requires 4096 MiB RAM; plan has 1024 MiB, 932 MiB detected. p95 ≤ 1000 ms · p99 ≤ 2000 ms · errors ≤ 1.0% ### Elasticsearch search and ingestion Insufficient RAM Full-text search, filtering and aggregations while new documents are indexed. Requires 4096 MiB RAM; plan has 1024 MiB, 932 MiB detected. p95 ≤ 250 ms · p99 ≤ 500 ms · errors ≤ 1.0% ### GitHub self-hosted runner Insufficient RAM A pinned build and test workflow with separate cold and warm dependency caches. Requires 8192 MiB RAM; plan has 1024 MiB, 932 MiB detected. p95 ≤ 3600000 ms · p99 ≤ 3600000 ms · errors ≤ 1.0% ### Odoo Community sales Insufficient RAM Customer lookup, sales order confirmation, stock operations and reporting. Requires 4096 MiB RAM; plan has 1024 MiB, 932 MiB detected. p95 ≤ 2000 ms · p99 ≤ 4000 ms · errors ≤ 1.0% ## Advertised vs. measured | Spec | Advertised | Measured | | --- | --- | --- | | vCPU | 1 | 1 | | CPU | Shared vCPU | 0% of busy CPU time stolen | | RAM | 1 GB | 932 MiB | | Storage | 20 GB NVMe | 18.7 GiB | | Architecture | x86\_64 | x86\_64 | | Virtualization | KVM | KVM | Guests usually report slightly less RAM and disk than advertised because of kernel reservations and filesystem overhead. ## Detected hardware **CPU model:** AMD EPYC-Genoa Processor **CPU cores:** 1 **CPU frequency:** 2396 MHz **AES-NI:** Yes **VM-x/AMD-V:** No **RAM reported:** 932 MiB **Swap reported:** 0 MiB **Disk reported:** 18.7 GiB **Virtualization reported:** KVM **CPU steal:** 0% of busy CPU time stolen **Distro:** Ubuntu 24.04.4 LTS **Kernel:** 6.8.0-138-generic **Architecture:** x86\_64 **IPv4 reachable:** Yes **IPv6 reachable:** Yes ## Advertised at test time **Plan:** CPX02 **Product line:** regular\_purpose **Location:** Singapore **vCPU:** 1 shared vCPU **Architecture:** x86\_64 **RAM:** 1 GB **Storage:** 20 GB NVMe **Virtualization:** KVM **Network:** Unknown **Transfer:** 21.99 TB **Port 25:** On request **Monthly price at test time:** $8.99 **Hourly price at test time:** $0.014 **YABS version:** v2026-09-20 **Ran at:** Oct 22, 2026, 04:19:11 UTC **Parser version:** 5 ## Run timeline **Queued:** Oct 21, 2026, 00:20:00 UTC **First update from the server:** Oct 22, 2026, 04:17:41 UTC **Completed:** Oct 22, 2026, 05:29:14 UTC **Server deleted:** Oct 22, 2026, 05:29:14 UTC **Cost of this run:** $0.017 Tests 15 of 16 done Live output 441 lines Setup 1 cores, 932.5 MiB RAM 4 of 4Done Disk (fio) 296 MB/s at 4k, 5.07 GB/s at 1m 5 of 5Done Network, IPv4 (iperf3) Median 850 Mbit/s send, 1.44 Gbit/s receive, 1 retried 3 of 3Done Web apps 1 failed, 9 skipped 2 of 3Some failed TestResultTime PostgreSQL transactions RuntimeError: TimeoutError: load generator worker did not respond Stopped RuntimeError: TimeoutError: load generator worker did not respond 10m 30s PostgreSQL large dataset Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. Not run Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. Redis cache 8 measured trials completed 8 trials Measurements, 8 measured trials completed 28m 33s Redis persistent queue 8 measured trials completed 8 trials Measurements, 8 measured trials completed 28m 23s Discourse community Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. Not run Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. n8n workflows Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. Not run Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. n8n queue workers Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. Not run Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. Ghost publishing Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. Not run Requires 2048 MiB RAM; plan has 1024 MiB, 932 MiB detected. Formbricks / Next.js Requires 4096 MiB RAM; plan has 1024 MiB, 932 MiB detected. Not run Requires 4096 MiB RAM; plan has 1024 MiB, 932 MiB detected. Elasticsearch search and ingestion Requires 4096 MiB RAM; plan has 1024 MiB, 932 MiB detected. Not run Requires 4096 MiB RAM; plan has 1024 MiB, 932 MiB detected. GitHub self-hosted runner Requires 8192 MiB RAM; plan has 1024 MiB, 932 MiB detected. Not run Requires 8192 MiB RAM; plan has 1024 MiB, 932 MiB detected. Odoo Community sales Requires 4096 MiB RAM; plan has 1024 MiB, 932 MiB detected. Not run Requires 4096 MiB RAM; plan has 1024 MiB, 932 MiB detected. Finish Results saved 1 of 1Done Watch party 0 watching This party closed at 06:29 UTC. It's read only now. ## Watch party Loading the party