bench(pk-default): add diagnostic config variants and correct the P3 analysis - #386
Merged
Conversation
…analysis Single-knob isolation (SHARPCOREDB_PK_DEFAULT_VARIANT=async|bufferedio|novalidate| noadaptive|hsinsert|plain|tuned) shows the earlier FullSync-dominance claim was wrong: WalDurabilityMode is only honored by GroupCommitWAL (off by default) and the async variant measures the same as default. NoEncryptMode moves default UPDATE ~1.3x (plain ~145K vs tuned ~109K median) and machine drift spans 275-315K on SQLite itself across the same window. docs/benchmarks/default-config-pk.md corrected: no blind 'FullSync commit-flush optimization'; the next step is a same-window interleaved A/B harness mode so default-vs-tuned deltas are attributable before any code change.
MPCoreDeveloper
force-pushed
the
fix/pk-default-analyse
branch
from
September 4, 2026 13:10
74b8f3c to
88ebf4e
Compare
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Samenvatting
P3-vervolg: FullSync-hypothese gefalsificeerd + diagnostische variant-harness.
In plaats van blind een “FullSync commit-flush optimalisatie” te implementeren (die nergens op
bleek te slaan), heb ik eerst geïsoleerd met median/single runs via nieuwe
SHARPCOREDB_PK_DEFAULT_VARIANT-configs:async(WalDurabilityMode)bufferedio/novalidatetuned(zónder NoEncryptMode)plain(= tuned + NoEncryptMode=true)Machine-drift is groot (SQLite zelf varieerde 275-315K in hetzelfde venster); deltas <~1.3x zijn
buiten een interleaved A/B niet toewijsbaar.
Actie:
docs/benchmarks/default-config-pk.mdgecorrigeerd (de eerdere “FullSync domineert”-claim stond er fout in) + harness-varianten toegevoegd als meetinstrument. Volgende stap is een
same-window interleaved A/B mode; pas daarna een optimalisatie op een gemeten knob.