chore: release v0.7.3

[sw-release-bot]

New release

coordinode 0.7.2 -> 0.7.3 (all 22 workspace crates share this version)

Added

• (operation) configurable admission window
• (operation) apply each client write once
• (embed) replicated embedded database behind a feature
• (storage) store index generations under local installations
• (migrate) add the node sizes report module
• (migrate) catalog record migrations, check and journal stats
• (migrate) offline tool to bring a data directory current

Fixed

• (storage) apply each recovered cache write once
• (storage) land a volatile write once, through the log
• (query) answer that an object exists from durable records
• (storage) never copy a checkpoint over a file it holds
• (storage) keep a node's own records across a rebuild
• (query) register a vector index generation once
• (query) report the outcome an index check recorded
• (query) register a text index generation once
• (raft) count blocking work holding the engine at shutdown
• (session) answer a commit repeated while it resolves
• (migrate) find every marked store, Raft snapshot captures too

Performance

• (storage) defer the log sync of volatile writes, not the write
• (storage) flush the device only where recovery needs it
• (storage) index registered commits by key
• (embed) publish nothing for an unchanged schema definition
• (query) bound property delta chains with whole writes

Testing

• (cluster) wait for a killed follower's replay
• (embed) measure how long a store takes to open
• (query) run the N+1 window tests on given instants
• (storage) cover crashes inside log truncation and compaction

Where the changes landed

The list above is the folded view. Each crate below carries the entries written
to its own changelog; a crate with no commits since the last release is absent.

coordinode-core

Added

• (operation) apply each client write once
• (embed) replicated embedded database behind a feature

Fixed

• (storage) apply each recovered cache write once
• (storage) land a volatile write once, through the log

Performance

• (storage) defer the log sync of volatile writes, not the write

coordinode-storage

Added

• (operation) apply each client write once
• (embed) replicated embedded database behind a feature

Fixed

• (storage) never copy a checkpoint over a file it holds
• (storage) apply each recovered cache write once
• (storage) land a volatile write once, through the log
• (storage) keep a node's own records across a rebuild

Performance

• (storage) flush the device only where recovery needs it
• (storage) defer the log sync of volatile writes, not the write
• (storage) index registered commits by key

Testing

• (storage) cover crashes inside log truncation and compaction

coordinode-cluster

Added

• (embed) replicated embedded database behind a feature

coordinode-modality

Added

• (embed) replicated embedded database behind a feature

Performance

• (storage) flush the device only where recovery needs it
• (storage) defer the log sync of volatile writes, not the write

coordinode-search

Performance

• (storage) defer the log sync of volatile writes, not the write

coordinode-query

Added

• (operation) apply each client write once
• (embed) replicated embedded database behind a feature

Fixed

• (query) answer that an object exists from durable records
• (storage) apply each recovered cache write once
• (query) report the outcome an index check recorded
• (query) register a vector index generation once
• (query) register a text index generation once

Performance

• (storage) flush the device only where recovery needs it
• (storage) defer the log sync of volatile writes, not the write

Testing

• (query) run the N+1 window tests on given instants

coordinode-test-fixtures

Performance

• (storage) flush the device only where recovery needs it

coordinode-raft

Added

• (operation) apply each client write once
• (embed) replicated embedded database behind a feature

Fixed

• (raft) count blocking work holding the engine at shutdown

Performance

• (storage) flush the device only where recovery needs it
• (storage) defer the log sync of volatile writes, not the write

coordinode-replicate

Added

• (embed) replicated embedded database behind a feature

Performance

• (storage) defer the log sync of volatile writes, not the write

coordinode-embed

Added

• (operation) configurable admission window
• (operation) apply each client write once
• (embed) replicated embedded database behind a feature

Fixed

• (query) answer that an object exists from durable records
• (storage) never copy a checkpoint over a file it holds
• (storage) apply each recovered cache write once
• (storage) land a volatile write once, through the log
• (storage) keep a node's own records across a rebuild
• (query) register a vector index generation once

Performance

• (storage) flush the device only where recovery needs it
• (storage) defer the log sync of volatile writes, not the write
• (storage) index registered commits by key
• (embed) publish nothing for an unchanged schema definition

Testing

• (embed) measure how long a store takes to open

coordinode-timeseries

Performance

• (storage) defer the log sync of volatile writes, not the write

coordinode-session

Added

• (operation) apply each client write once

Fixed

• (session) answer a commit repeated while it resolves

coordinode-client

Added

• (operation) apply each client write once

coordinode-migrate

Added

• (operation) apply each client write once
• (storage) store index generations under local installations
• (migrate) add the node sizes report module
• (migrate) catalog record migrations, check and journal stats
• (migrate) offline tool to bring a data directory current

Fixed

• (storage) never copy a checkpoint over a file it holds
• (migrate) find every marked store, Raft snapshot captures too

Performance

• (query) bound property delta chains with whole writes

---

This PR was generated with release-plz.

[coderabbitai]

Important

Review skipped

Bot user detected.

To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

• Configuration used: Organization UI
• Review profile: ASSERTIVE
• Plan: Advanced
• Run ID: 4f7f1544-f772-4147-be44-5d41907cce3f

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}