Bench run: dcompact / 31298143542

generated (UTC): 2026-08-09T09:29:35.567342+00:00

ToplingDB zipkeyonly db_bench | ToplingDB zipkeyvalue db_bench | RocksDB v8.10 db_bench | RocksDB master (5fe22535) db_bench | DB INFO LOGs (Actions artifact) — ToplingDB zipkeyonly LOG-fillrandom, ToplingDB zipkeyonly LOG-fillseq, ToplingDB zipkeyvalue LOG-fillrandom, ToplingDB zipkeyvalue LOG-fillseq, RocksDB v8.10 LOG-fillrandom, RocksDB v8.10 LOG-fillseq, RocksDB master (5fe22535) LOG-fillrandom, RocksDB master (5fe22535) LOG-fillseq

ToplingDB bench yaml (runtime graft): ToplingDB zipkeyonly db_bench-fillrandom.yaml | ToplingDB zipkeyonly db_bench-fillseq.yaml | ToplingDB zipkeyvalue db_bench-fillrandom.yaml | ToplingDB zipkeyvalue db_bench-fillseq.yaml

Bench configuration (dcompact)

Write-side db_bench CPU quota: 50% (from runner_env.txt cpu_quota_write). Topling compact: dcompact_worker — dcompact_worker runs outside the write cgroup and uses remaining CPU cores. RocksDB compact: compaction_service_spool — CompactionService spool + out-of-cgroup broker/worker on remaining cores.

Runner hardware/software

OS	Ubuntu 24.04.4 LTS
Kernel	6.17.0-1020-azure
CPU	AMD EPYC 9V74 80-Core Processor
Logical CPUs	4
Memory	15.6GiB
/dev/shm	7.8GiB
RocksDB block cache	7.8GiB

On-disk DB size (3.7GiB) ≤ block cache (7.8GiB) — cache can hold the entire dataset. So this benchmark mainly shows CPU and memory cost.

/dev/shm usage (disk space; after db_bench)

Allocated disk usage (IEC blocks). RocksDB uses default Snappy compression. Space ratio = engine / v8.10; <1 = less space than RocksDB, >1 = larger.

workload	ToplingDB zipkeyonly	ToplingDB zipkeyvalue	RocksDB v8.10	RocksDB master (5fe22535)	zipkeyonly / v8.10 (space)	zipkeyvalue / v8.10 (space)
fillrandom	3.0GiB	494.2MiB	3.7GiB	3.7GiB	0.81x	0.13x
fillseq	4.7GiB	1.1GiB	3.2GiB	3.2GiB	1.45x	0.35x

Peak RSS (RAM; during db_bench)

RSS is Resident Set Size. RocksDB block cache = half physical memory (7.8GiB). Ratio = engine / v8.10; <1 = less RSS, >1 = more RSS.

workload	ToplingDB zipkeyonly	ToplingDB zipkeyvalue	RocksDB v8.10	RocksDB master (5fe22535)	zipkeyonly / v8.10 (RSS)	zipkeyvalue / v8.10 (RSS)
fillrandom	788.9MiB	852.2MiB	4.1GiB	4.2GiB	0.19x	0.20x
fillseq	513.3MiB	1.2GiB	6.4GiB	6.5GiB	0.08x	0.19x

RAM usage over time

Sampled once per second from /proc/statm plus open-file page cache. Colored bands show benchmark segments (start time from db_bench output).

rss: resident set size (pages currently in RAM for the process); rss = shared + anony.
shared: shared resident pages; mostly readonly (cheap; OS prefers reclaiming these, no swap needed).
anony: rss - shared; mostly readwrite anonymous pages (costly, needs swap).
pagecache: kernel file page cache for regular files the process currently has open (cachestat(2) on /proc/pid/fd, deduped by inode; covers buffered readwrite and mmap). Pages brought in only via buffered I/O are not charged to process RSS; mmap'd file pages can appear in both pagecache and RSS/shared, so series may overlap.
anony+pc: anony+pagecache (sum of process anonymous RSS and open-file page cache; not a disjoint partition).

Comparison: db_bench fillrandom suite (perf)

Benchmarks: fillrandom, flush, compact, readseq×3, readrandom. RocksDB uses default Snappy compression. compact row shows operations/seconds. RocksDB time / zipkey*: >1 = that zipkey* engine faster. Values show ops/sec.

benchmark	ToplingDB zipkeyonly ops/sec	ToplingDB zipkeyvalue ops/sec	RocksDB v8.10 ops/sec	RocksDB master (5fe22535) ops/sec	zipkeyvalue time / zipkeyonly	v8.10 time / zipkeyonly	master (5fe22535) time / zipkeyonly	v8.10 time / zipkeyvalue	master (5fe22535) time / zipkeyvalue
compact	1/53.2s	1/79.4s	1/49.8s	1/42.3s	1.49x	0.94x	0.80x	0.63x	0.53x
fillrandom	243251	242659	218511	215282	1.00x	1.11x	1.13x	1.11x	1.13x
readrandom	1805200	732967	146945	138751	2.46x	12.28x	13.01x	4.99x	5.28x
readseq	11375101	8347565	4509772	3793923	1.36x	1.59x	1.90x	1.17x	1.39x

Comparison: db_bench fillseq suite (perf)

Same as fillrandom. Watch compact / readseq / readrandom cost for minDictZip=10 vs space savings above.

benchmark	ToplingDB zipkeyonly ops/sec	ToplingDB zipkeyvalue ops/sec	RocksDB v8.10 ops/sec	RocksDB master (5fe22535) ops/sec	zipkeyvalue time / zipkeyonly	v8.10 time / zipkeyonly	master (5fe22535) time / zipkeyonly	v8.10 time / zipkeyvalue	master (5fe22535) time / zipkeyvalue
compact	1/83.7s	1/133.7s	1/39.8s	1/37.1s	1.60x	0.48x	0.44x	0.30x	0.28x
fillseq	581853	574490	954671	922932	1.01x	0.61x	0.63x	0.60x	0.62x
readrandom	2795221	581669	140829	136930	4.81x	19.85x	20.41x	4.13x	4.25x
readseq	13153138	9932354	4568946	3873774	1.32x	2.88x	3.40x	2.17x	2.56x

Per-engine details

ToplingDB zipkeyonly

db_bench (fillrandom+compact+readseq/readrandom)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillrandom	4.110974	243251	411.098	100000000	13.5 MB/s
compact	53234221.000000	0	53.234	1
readseq	0.087752	11395723	8.775	100000000	630.3 MB/s
readseq	0.088181	11340277	8.818	100000000	627.3 MB/s
readseq	0.087911	11375101	8.791	100000000	629.2 MB/s
readrandom	0.553954	1805200	55.395	100000000	63.1 MB/s (63205377 of 100000000 found)

db_bench (fillseq suite)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillseq	1.718646	581853	171.865	100000000	32.2 MB/s
compact	83696267.000000	0	83.696	1
readseq	0.076121	13136989	7.612	100000000	726.6 MB/s
readseq	0.075070	13320804	7.507	100000000	736.8 MB/s
readseq	0.076027	13153138	7.603	100000000	727.5 MB/s
readrandom	0.357753	2795221	35.775	100000000	154.6 MB/s (100000000 of 100000000 found)

ToplingDB zipkeyvalue

db_bench (fillrandom+compact+readseq/readrandom)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillrandom	4.121005	242659	412.101	100000000	13.4 MB/s
compact	79417715.000000	0	79.418	1
readseq	0.121148	8254361	12.115	100000000	456.6 MB/s
readseq	0.119826	8345393	11.983	100000000	461.6 MB/s
readseq	0.119795	8347565	11.980	100000000	461.7 MB/s
readrandom	1.364317	732967	136.432	100000000	25.6 MB/s (63206447 of 100000000 found)

db_bench (fillseq suite)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillseq	1.740671	574490	174.067	100000000	31.8 MB/s
compact	133732789.000000	0	133.733	1
readseq	0.103022	9706662	10.302	100000000	536.9 MB/s
readseq	0.100957	9905136	10.096	100000000	547.9 MB/s
readseq	0.100681	9932354	10.068	100000000	549.4 MB/s
readrandom	1.719190	581669	171.919	100000000	32.2 MB/s (100000000 of 100000000 found)

RocksDB v8.10

db_bench (fillrandom+compact+readseq/readrandom)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillrandom	4.576	218511	457.641	100000000	12.1 MB/s
compact	49783636.000	0	49.784	1
readseq	0.325	3080731	20.518	63209518	170.4 MB/s
readseq	0.221	4519915	13.985	63209518	250.0 MB/s
readseq	0.222	4509772	14.016	63209518	249.4 MB/s
readrandom	6.805	146945	680.523	100000000	5.1 MB/s (63211078 of 100000000 found)

db_bench (fillseq suite)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillseq	1.047	954671	104.748	100000000	52.8 MB/s
compact	39795711.000	0	39.796	1
readseq	0.381	2624067	38.109	100000000	145.1 MB/s
readseq	0.222	4499214	22.226	100000000	248.9 MB/s
readseq	0.219	4568946	21.887	100000000	252.7 MB/s
readrandom	7.101	140829	710.077	100000000	7.8 MB/s (100000000 of 100000000 found)

RocksDB master (5fe22535)

db_bench (fillrandom+compact+readseq/readrandom)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillrandom	4.645	215282	464.506	100000000	11.9 MB/s
compact	42323790.000	0	42.324	1
readseq	0.386	2593059	24.378	63213751	143.4 MB/s
readseq	0.261	3824398	16.529	63213751	211.5 MB/s
readseq	0.264	3793923	16.662	63213751	209.9 MB/s
readrandom	7.207	138751	720.713	100000000	4.9 MB/s (63215063 of 100000000 found)

db_bench (fillseq suite)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillseq	1.084	922932	108.350	100000000	51.1 MB/s
compact	37117514.000	0	37.118	1
readseq	0.427	2343664	42.668	100000000	129.6 MB/s
readseq	0.259	3864371	25.877	100000000	213.8 MB/s
readseq	0.258	3873774	25.815	100000000	214.3 MB/s
readrandom	7.303	136930	730.300	100000000	7.6 MB/s (100000000 of 100000000 found)