Skip to content

Docs: a failed load can keep different rows on the two packages (closes #174) - #212

Merged
tae898 merged 1 commit into
mainfrom
docs/174-commit-default-per-package
Nov 3, 2026
Merged

tae898 merged 1 commit into
mainfrom
docs/174-commit-default-per-package

Conversation

@tae898

@tae898 tae898 commented Nov 3, 2026

Copy link
Copy Markdown
Collaborator

The commit-size defaults of insert_many, graph_batch and create_vertices are measured per build (native 1000, JVM 10000, the A/B is on the bulk-loads page) and stay as they are. The page already explains what a failed load keeps; this adds one paragraph saying that the same program can therefore keep different rows on the two packages after the same failure, and how to make the outcome identical (the same commit_every on both, or commit_every=0). No behaviour change.

Closes #174

🤖 Generated with Claude Code

…ges (closes #174)

The commit-size defaults are measured per build (native 1000, JVM 10000) and stay; the bulk-loads page already explained what a failed load keeps. This adds the cross-package consequence and the way to make it identical (the same commit_every on both, or commit_every=0). No behaviour change.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@tae898
tae898 merged commit 1b315c1 into main Nov 3, 2026
1 check passed
@tae898
tae898 deleted the docs/174-commit-default-per-package branch November 3, 2026 15:23
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

insert_many: native and JVM keep different data after a failed load (commit_every defaults 1,000 vs 10,000)

1 participant