Repository navigation
Docs: say when to turn on the JVM AOT cache, with measured figures - #248
Merged
Merged
Conversation
…loses #247) The timed A/B kept the AOT cache opt-in: the first script is faster (0.653, 95% interval [0.648, 0.658]) but no steady figure gains, and the 300k-row insert interval ended at 1.036 against a 1.03 bar. By the study's rule the cache is documented, not made the default. Figures are generated from the study's score.out by script. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
…ng run Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Closes #247
Adds a section to the installation page, "Faster start-up with the AOT cache" (
#jvm-aot), and links it from the start-up row ofwhich-package.adocand from the container list. It names the opt-in as the code has it (ARCADEDB_AOT=1orstart_jvm(aot=True), withARCADEDB_AOT_CACHE=<file>for the file), says what the cache does, when to turn it on (short-lived processes where start-up dominates) and when not to bother (long-running services), the one-time costs (a training run, a check run and a cache file of about 40 MiB), and the measured figures with 95% intervals, stated as one laptop study on JDK 25 with the study's workload. It makes no speed promise beyond those figures.The cache stays off by default because the timed study's pre-written rule was not met: the first script is faster with the cache, but the 300,000-row
insert_manyinterval ended at 1.036 against a bar of 1.03, and no steady figure showed a gain we can rely on.The figures, the pair and round counts and the one-time costs are written into the page by a script that reads the study's score output; none is typed by hand. The script also asserts the claims the prose makes about the intervals (the control arm includes 1, the 2-hop interval ends below 1, the others include 1). Docs only: no change in
srcortests. TheDocsworkflow (Antora, a warning fails it) is the check for this pull request.🤖 Generated with Claude Code