Bench run: dcompact / 31303589765

generated (UTC): 2026-08-09T10:21:07.496544+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: 150% (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 7763 64-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	493.2MiB	3.7GiB	3.7GiB	0.81x	0.13x
fillseq	4.7GiB	1.2GiB	3.2GiB	3.2GiB	1.46x	0.36x

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	858.6MiB	827.1MiB	4.1GiB	4.2GiB	0.21x	0.20x
fillseq	511.3MiB	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	1/50.5s	1/81.6s	1/59.5s	1/61.9s	1.62x	1.18x	1.22x	0.73x	0.76x
fillrandom	622466	631838	488920	553818	0.99x	1.27x	1.12x	1.29x	1.14x
readrandom	3788691	1610986	329753	328572	2.35x	11.49x	11.53x	4.89x	4.90x
readseq	22386604	16622249	10692768	8033999	1.35x	1.32x	1.76x	0.98x	1.31x

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	1/88.5s	1/130.5s	1/40.5s	1/41.0s	1.47x	0.46x	0.46x	0.31x	0.31x
fillseq	1765245	1773186	2888375	2735159	1.00x	0.61x	0.65x	0.61x	0.65x
readrandom	6603202	1351010	321635	353332	4.89x	20.53x	18.69x	4.20x	3.82x
readseq	25695592	20736270	10750388	8022683	1.24x	2.39x	3.20x	1.93x	2.59x

Per-engine details

ToplingDB zipkeyonly

db_bench (fillrandom+compact+readseq/readrandom)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillrandom	1.606510	622466	160.651	100000000	34.4 MB/s
compact	50513202.000000	0	50.513	1
readseq	0.044497	22472930	4.450	100000000	1243.0 MB/s
readseq	0.044640	22400972	4.464	100000000	1239.1 MB/s
readseq	0.044669	22386604	4.467	100000000	1238.3 MB/s
readrandom	0.263943	3788691	26.394	100000000	132.5 MB/s (63217682 of 100000000 found)

db_bench (fillseq suite)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillseq	0.566493	1765245	56.649	100000000	97.6 MB/s
compact	88480306.000000	0	88.480	1
readseq	0.037855	26416666	3.785	100000000	1461.2 MB/s
readseq	0.037682	26538010	3.768	100000000	1467.9 MB/s
readseq	0.038917	25695592	3.892	100000000	1421.3 MB/s
readrandom	0.151441	6603202	15.144	100000000	365.2 MB/s (100000000 of 100000000 found)

ToplingDB zipkeyvalue

db_bench (fillrandom+compact+readseq/readrandom)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillrandom	1.582684	631838	158.268	100000000	34.9 MB/s
compact	81590536.000000	0	81.591	1
readseq	0.059727	16742627	5.973	100000000	926.1 MB/s
readseq	0.059636	16768243	5.964	100000000	927.5 MB/s
readseq	0.060160	16622249	6.016	100000000	919.4 MB/s
readrandom	0.620737	1610986	62.074	100000000	56.3 MB/s (63211220 of 100000000 found)

db_bench (fillseq suite)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillseq	0.563956	1773186	56.396	100000000	98.1 MB/s
compact	130502198.000000	0	130.502	1
readseq	0.049035	20393525	4.904	100000000	1128.0 MB/s
readseq	0.048921	20440877	4.892	100000000	1130.6 MB/s
readseq	0.048224	20736270	4.822	100000000	1147.0 MB/s
readrandom	0.740186	1351010	74.019	100000000	74.7 MB/s (100000000 of 100000000 found)

RocksDB v8.10

db_bench (fillrandom+compact+readseq/readrandom)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillrandom	2.045	488920	204.532	100000000	27.0 MB/s
compact	59514511.000	0	59.515	1
readseq	0.134	7483133	8.447	63213445	413.9 MB/s
readseq	0.094	10680755	5.918	63213445	590.8 MB/s
readseq	0.094	10692768	5.912	63213445	591.5 MB/s
readrandom	3.033	329753	303.257	100000000	11.5 MB/s (63220099 of 100000000 found)

db_bench (fillseq suite)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillseq	0.346	2888375	34.622	100000000	159.8 MB/s
compact	40461243.000	0	40.461	1
readseq	0.160	6260564	15.973	100000000	346.3 MB/s
readseq	0.094	10688055	9.356	100000000	591.2 MB/s
readseq	0.093	10750388	9.302	100000000	594.6 MB/s
readrandom	3.109	321635	310.911	100000000	17.8 MB/s (100000000 of 100000000 found)

RocksDB master (5fe22535)

db_bench (fillrandom+compact+readseq/readrandom)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillrandom	1.806	553818	180.565	100000000	30.6 MB/s
compact	61850972.000	0	61.851	1
readseq	0.173	5786482	10.925	63214451	320.1 MB/s
readseq	0.125	7968744	7.933	63214451	440.8 MB/s
readseq	0.124	8033999	7.868	63214451	444.4 MB/s
readrandom	3.043	328572	304.346	100000000	11.5 MB/s (63205195 of 100000000 found)

db_bench (fillseq suite)

benchmark	micros/op	ops/sec	seconds	operations	extra
fillseq	0.366	2735159	36.561	100000000	151.3 MB/s
compact	40977270.000	0	40.977	1
readseq	0.193	5168474	19.348	100000000	285.9 MB/s
readseq	0.124	8040786	12.437	100000000	444.8 MB/s
readseq	0.125	8022683	12.465	100000000	443.8 MB/s
readrandom	2.830	353332	283.020	100000000	19.5 MB/s (100000000 of 100000000 found)