ovh / waw — vps-1-lz-2026

€7.49/mo · storage net-fast · host 99ba7e30ec33 · 2026-07-17T15:22:22Z · submitter uptimeify

Profiles

profile	grade	bound by
postgres_oltp	C	`disk.rand_read_8k.iops`
timescale_ingest	C	`disk.rand_write_8k.iops`
patroni_member	F	`cpu.single_thread_eps`
redis_sentinel	F	`cpu.single_thread_eps`
worker_probe	F	`cpu.single_thread_eps`
playwright_node	F	`cpu.single_thread_eps`
nuxt_ssr	F	`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 3 ms B

−"NVMe SSD — low latency"

+ p99.9 vs median, not the mean p50 0.15 ms → p99.9 0.58 ms A

−"Dedicated vCPUs"

+ CPU steal measured under full load 0.02% steal

−"High-performance cores"

+ single-thread throughput 350 eps F

−"Fast" (short benchmark)

+ IOPS drop over 30 min sustained load 0.27% degradation

−"No noisy neighbours"

+ scheduler stall p99.9 1.13 ms C

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

0.62 ms

p50

0.59 ms

p99

1.1 ms

p99.9

3 ms B

max

16.3 ms

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

read QD1

0.58 ms

read

9.2 ms

write

38.5 ms

mixed

8.4 ms

Throughput

seq read

3,535 MB/s

seq write

2,593 MB/s

Random IOPS

read 8K

20,031 IOPS

write 8K

17,095 IOPS

mixed 8K

20,033 IOPS

fsync

1,462 IOPS

Sustained IOPS — first vs last minute of 30 min 0.27% degradation over 30 min

first min

28,675 IOPS

last min

28,597 IOPS

exact values & per-metric grades

metric	value	grade
`disk.wal_fsync.p999_us`	2998.27	B
`disk.wal_fsync.iops`	1461.86	B
`disk.rand_read_8k_qd1.p99_us`	238.59	A
`disk.rand_read_8k.iops`	20031.26	C
`disk.rand_write_8k.iops`	17094.62	C
`disk.seq_write.bw_mbs`	2592.92	A
`disk.seq_read.bw_mbs`	3534.91	A
`disk.steady_state.degradation_pct`	0.27	A

CPU

bound by cpu.single_thread_eps

Throughput (events/s)

single

350 eps F

multi

1,308 eps

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

p99

0.13 ms

p99.9

1.1 ms C

max

8.4 ms

Sustained throughput — first vs last minute -2% degradation · steal 0.03%

first min

1,318 eps

last min

1,354 eps

exact values & per-metric grades

metric	value	grade
`cpu.single_thread_eps`	349.98	F
`cpu.scaling_efficiency`	0.934	A
`cpu.steal_pct_under_load`	0.02	A
`cpu.stall_p999_us`	1126	C
`cpu.steady_state.degradation_pct`	-2	A
`cpu.tls_verify_s`	9828.5	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

24,844 MB/s C

seq read

56,024 MB/s

seq write

39,607 MB/s

rnd read

6,524 MB/s

rnd write

663 MB/s

exact values & per-metric grades

metric	value	grade
`ram.bw_read_mbs`	24844.41	C

Network

bound by network.loss_pct

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

hetzner-fsn1

24.2 ms

hetzner-hel1

41.4 ms

hetzner-ash

104 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`	1.0083333333333333	A