upcloud / de-fra1 — STARTER-4xCPU-8GB

€20/mo · storage net-fast · host cbfac87a9740 · 2026-07-19T12:52:32Z · submitter uptimeify

Profiles

profile	grade	bound by
postgres_oltp	D	`disk.rand_read_8k.iops`
timescale_ingest	D	`disk.seq_read.bw_mbs`
patroni_member	D	`cpu.single_thread_eps`
redis_sentinel	D	`cpu.single_thread_eps`
worker_probe	D	`cpu.single_thread_eps`
playwright_node	D	`cpu.single_thread_eps`
nuxt_ssr	D	`cpu.single_thread_eps`

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 4.62 ms C

−"NVMe SSD — low latency"

+ p99.9 vs median, not the mean p50 0.1 ms → p99.9 0.87 ms A

−"Dedicated vCPUs"

+ CPU steal measured under full load 0.05% steal

−"High-performance cores"

+ single-thread throughput 629 eps D

−"Fast" (short benchmark)

+ IOPS drop over 30 min sustained load 0% degradation

−"No noisy neighbours"

+ scheduler stall p99.9 0.21 ms B

Disk

bound by disk.rand_read_8k.iops

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

0.79 ms

p50

0.69 ms

p99

2.2 ms

p99.9

4.6 ms C

max

25.9 ms

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

read QD1

0.87 ms

read

98 ms

write

2,139 ms

mixed

90.7 ms

Throughput

seq read

400 MB/s

seq write

400 MB/s

Random IOPS

read 8K

10,002 IOPS

write 8K

5,312 IOPS

mixed 8K

9,504 IOPS

fsync

1,206 IOPS

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

first min

12,805 IOPS

last min

14,281 IOPS

exact values & per-metric grades

metric	value	grade
`disk.wal_fsync.p999_us`	4620.29	C
`disk.wal_fsync.iops`	1205.71	B
`disk.rand_read_8k_qd1.p99_us`	309.25	A
`disk.rand_read_8k.iops`	10002.27	D
`disk.rand_write_8k.iops`	5311.77	D
`disk.seq_write.bw_mbs`	400.33	C
`disk.seq_read.bw_mbs`	400.44	D
`disk.steady_state.degradation_pct`	0	A

CPU

bound by cpu.single_thread_eps

Throughput (events/s)

single

629 eps D

multi

2,508 eps

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

p99

0.07 ms

p99.9

0.21 ms B

max

1.4 ms

Sustained throughput — first vs last minute 0% degradation · steal 0.07%

first min

2,504 eps

last min

2,512 eps

exact values & per-metric grades

metric	value	grade
`cpu.single_thread_eps`	629.47	D
`cpu.scaling_efficiency`	0.996	A
`cpu.steal_pct_under_load`	0.05	A
`cpu.stall_p999_us`	209	B
`cpu.steady_state.degradation_pct`	0	A
`cpu.tls_verify_s`	14477.3	C

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

51,341 MB/s A

seq read

184,914 MB/s

seq write

31,737 MB/s

rnd read

11,797 MB/s

rnd write

1,002 MB/s

exact values & per-metric grades

metric	value	grade
`ram.bw_read_mbs`	51340.92	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

5.9 ms

hetzner-hel1

23.9 ms

hetzner-ash

93.9 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`	1.148936170212766	B