hetzner / hel-1 — CPX32

€42.23/mo · storage net-fast · host ffe9c660efdd · 2026-07-17T15:22:56Z · submitter uptimeify

Profiles

profile	grade	bound by
postgres_oltp	C	`disk.wal_fsync.p999_us`
timescale_ingest	C	`disk.wal_fsync.p999_us`
patroni_member	C	`disk.wal_fsync.p999_us`
redis_sentinel	C	`disk.wal_fsync.p999_us`
worker_probe	B	`cpu.stall_p999_us`
playwright_node	C	`host.ram_mb`
nuxt_ssr	B	`cpu.stall_p999_us`

Tail-latency reality

fsync cumulative distribution

hover the curve — snaps to measured percentiles

Points are the measured p50 / p99 / p99.9 / max — the line between them is interpolated (we store percentiles, not a full histogram). Fixed axis: a tight curve on the left is a good host; a curve swept right is an exploding tail.

marketing claim vs measured

− typical marketing claim + what p99bench measured here

−"Up to N,000 IOPS"

+ fsync p99.9 latency at QD1 3.69 ms C

−"NVMe SSD — low latency"

+ p99.9 vs median, not the mean p50 0.15 ms → p99.9 0.87 ms A

−"Dedicated vCPUs"

+ CPU steal measured under full load 0% steal

−"High-performance cores"

+ single-thread throughput 1,658 eps A

−"Fast" (short benchmark)

+ IOPS drop over 30 min sustained load 0% degradation

−"No noisy neighbours"

+ scheduler stall p99.9 0.25 ms B

Disk

bound by disk.wal_fsync.p999_us

WAL fsync latency ladder (ms, QD1) Every COMMIT waits on one fdatasync. p99.9 is the graded metric — the tail that stalls a database.

avg

0.67 ms

p50

0.65 ms

p99

1.1 ms

p99.9

3.7 ms C

max

28.6 ms

8K random latency p99.9 by operation (ms)

read QD1

0.87 ms

read

3.7 ms

write

14.6 ms

mixed

4.8 ms

Throughput

seq read

6,885 MB/s

seq write

5,440 MB/s

Random IOPS

read 8K

80,576 IOPS

write 8K

66,085 IOPS

mixed 8K

50,547 IOPS

fsync

1,351 IOPS

Sustained IOPS — first vs last minute of 30 min 0% degradation over 30 min

first min

71,349 IOPS

last min

73,853 IOPS

exact values & per-metric grades

metric	value	grade
`disk.wal_fsync.p999_us`	3686.4	C
`disk.wal_fsync.iops`	1350.78	B
`disk.rand_read_8k_qd1.p99_us`	329.73	A
`disk.rand_read_8k.iops`	80576.12	B
`disk.rand_write_8k.iops`	66084.51	A
`disk.seq_write.bw_mbs`	5440.03	A
`disk.seq_read.bw_mbs`	6884.98	A
`disk.steady_state.degradation_pct`	0	A

CPU

bound by cpu.stall_p999_us

Throughput (events/s)

single

1,658 eps A

multi

6,537 eps

Scheduler stall ladder (ms) How long the kernel keeps a runnable thread waiting. p99.9 is graded.

p99

0.14 ms

p99.9

0.25 ms B

max

3.9 ms

Sustained throughput — first vs last minute 0% degradation · steal 0%

first min

6,536 eps

last min

6,471 eps

exact values & per-metric grades

metric	value	grade
`cpu.single_thread_eps`	1658.01	A
`cpu.scaling_efficiency`	0.985	A
`cpu.steal_pct_under_load`	0	A
`cpu.stall_p999_us`	249	B
`cpu.steady_state.degradation_pct`	0	A
`cpu.tls_verify_s`	16767	B

RAM

bound by ram.bw_read_mbs

Memory bandwidth (MB/s) bw read is measured outside cache (the graded metric); seq/rnd are raw STREAM-style figures.

bw read

57,350 MB/s A

seq read

210,696 MB/s

seq write

16,127 MB/s

rnd read

15,827 MB/s

rnd write

958 MB/s

exact values & per-metric grades

metric	value	grade
`ram.bw_read_mbs`	57350.3	A

Network

bound by network.rtt_jitter_ratio

Round-trip p99 by target (ms) Same fixed target list from every host — distance is constant, so peering and neighbour noise surface.

hetzner-fsn1

25.4 ms

hetzner-hel1

0.58 ms

hetzner-ash

111 ms

Packet loss by target (%)

hetzner-fsn1

0 %

hetzner-hel1

0 %

hetzner-ash

0 %

exact values & per-metric grades

metric	value	grade
`network.loss_pct`	0	A
`network.rtt_jitter_ratio`	1.2653508771929822	B