Skip to content

feat(table): support lookup changelog production #1252

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
JingsongLi merged 2 commits into from
Oct 29, 2026
Merged

feat(table): support lookup changelog production #1252

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
e453770
feat(table): add lookup changelog production for primary-key compaction
JingsongLi Oct 29, 2026
c7233be
fix(table): align lookup changelog with Java reducers
JingsongLi Oct 29, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 4 additions & 0 deletions crates/paimon/src/io/file_io.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -584,6 +584,10 @@ impl FileIO {
Ok(self.cache_namespace.to_string())
}

pub(crate) fn local_cache(&self) -> Option<Arc<LocalCache>> {
self.cache.clone()
}

#[cfg(test)]
pub(crate) fn has_local_cache(&self) -> bool {
self.cache.is_some()
Expand Down
102 changes: 98 additions & 4 deletions crates/paimon/src/table/compact.rs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -111,6 +111,20 @@ impl<'a> CompactBuilder<'a> {
/// Executor resources are local; input files and output settings come from each task.
pub fn new_executor(&self) -> Result<CompactExecutor> {
ensure_supported(self.table)?;
if self
.table
.schema()
.core_options()
.try_changelog_producer()?
== ChangelogProducer::Lookup
&& self.table.file_io().local_cache().is_none()
{
return Err(Error::ConfigInvalid {
message:
"Lookup compaction requires local-cache.enabled=true on the catalog/FileIO"
.into(),
});
}
Ok(CompactExecutor {
table: self.table.clone(),
resources: self.resources.clone(),
Expand Down Expand Up @@ -274,19 +288,36 @@ pub struct CompactExecutor {
pub struct CompactResult {
task: CompactTask,
files: Vec<DataFileMeta>,
#[serde(default)]
changelog_files: Vec<DataFileMeta>,
}

impl CompactResult {
pub fn files(&self) -> &[DataFileMeta] {
&self.files
}

pub fn changelog_files(&self) -> &[DataFileMeta] {
&self.changelog_files
}

pub fn metrics(&self) -> CompactMetrics {
CompactMetrics {
files_removed: self.task.files.len(),
files_added: self.files.len(),
rows_written: self.files.iter().map(|file| file.row_count as u64).sum(),
bytes_written: self.files.iter().map(|file| file.file_size as u64).sum(),
changelog_files_added: self.changelog_files.len(),
changelog_rows_written: self
.changelog_files
.iter()
.map(|file| file.row_count as u64)
.sum(),
changelog_bytes_written: self
.changelog_files
.iter()
.map(|file| file.file_size as u64)
.sum(),
}
}
}
Expand All @@ -298,6 +329,12 @@ pub struct CompactMetrics {
pub files_added: usize,
pub rows_written: u64,
pub bytes_written: u64,
#[serde(default)]
pub changelog_files_added: usize,
#[serde(default)]
pub changelog_rows_written: u64,
#[serde(default)]
pub changelog_bytes_written: u64,
}

impl CompactExecutor {
Expand Down Expand Up @@ -351,13 +388,16 @@ impl CompactExecutor {
return Ok(CompactResult {
task: task.clone(),
files,
changelog_files: vec![],
});
}
if task.is_primary_key() {
let files = primary_key::execute(&table, task, &split, self.resources.clone()).await?;
let (files, changelog_files) =
primary_key::execute(&table, task, &split, self.resources.clone()).await?;
return Ok(CompactResult {
task: task.clone(),
files,
changelog_files,
});
}
let mut read = table.new_read_builder().new_read()?.with_ordered(true);
Expand Down Expand Up @@ -408,6 +448,7 @@ impl CompactExecutor {
Ok(files) => Ok(CompactResult {
task: task.clone(),
files,
changelog_files: vec![],
}),
Err(error) => {
writer.abort().await;
Expand Down Expand Up @@ -513,6 +554,7 @@ impl CompactTask {
/// Validated inputs for the existing atomic snapshot publication machinery.
pub(super) struct PreparedCompact {
pub entries: Vec<ManifestEntry>,
pub changelog_entries: Vec<ManifestEntry>,
pub schema_id: i64,
pub snapshot_id: i64,
pub read_snapshot: Option<Box<Snapshot>>,
Expand All @@ -529,6 +571,7 @@ pub(super) fn prepare_results(table: &Table, results: &[CompactResult]) -> Resul
let mut inputs = HashSet::new();
let mut outputs = HashSet::new();
let mut entries = Vec::new();
let mut changelog_entries = Vec::new();
let mut metrics = CompactMetrics::default();
for result in results {
let task = &result.task;
Expand Down Expand Up @@ -605,11 +648,55 @@ pub(super) fn prepare_results(table: &Table, results: &[CompactResult]) -> Resul
"PK compaction must retain each key's merged state unless retracts are ignored",
));
}
let lookup = options.try_changelog_producer()? == ChangelogProducer::Lookup;
if !lookup && !result.changelog_files.is_empty() {
return Err(invalid(
"Only lookup compaction may produce changelog files",
));
}
let mut changelog_rows = 0u64;
for file in &result.changelog_files {
validate_file(file, true)?;
if file.schema_id != task.context.schema.id()
|| file.file_source != Some(1)
|| file.level != task.output_level()
|| file
.delete_row_count
.is_none_or(|count| count > file.row_count)
|| !outputs.insert(file.file_name.clone())
|| task.files.iter().any(|old| old.file_name == file.file_name)
{
return Err(invalid("Invalid compaction changelog file"));
}
changelog_rows += file.row_count as u64;
changelog_entries.push(ManifestEntry::new(
FileKind::Add,
task.partition.clone(),
task.bucket,
task.total_buckets,
file.clone(),
2,
));
}
let l0_rows = task
.files
.iter()
.filter(|file| file.level == 0)
.map(|file| file.row_count as u64)
.sum::<u64>();
if changelog_rows > l0_rows.saturating_mul(2) {
return Err(invalid(
"Lookup changelogs must be generated by level-0 records",
));
}
let part = result.metrics();
metrics.files_removed += part.files_removed;
metrics.files_added += part.files_added;
metrics.rows_written += part.rows_written;
metrics.bytes_written += part.bytes_written;
metrics.changelog_files_added += part.changelog_files_added;
metrics.changelog_rows_written += part.changelog_rows_written;
metrics.changelog_bytes_written += part.changelog_bytes_written;
}
ids.sort_unstable();
if ids.windows(2).any(|pair| pair[0] == pair[1]) {
Expand All @@ -623,6 +710,7 @@ pub(super) fn prepare_results(table: &Table, results: &[CompactResult]) -> Resul
}
Ok(PreparedCompact {
entries,
changelog_entries,
schema_id: first.task.context.schema.id(),
snapshot_id: first.task.snapshot_id(),
read_snapshot: first
Expand Down Expand Up @@ -692,15 +780,21 @@ pub(super) fn ensure_supported(table: &Table) -> Result<()> {
&& (options.bucket() == 0
|| !matches!(
options.try_changelog_producer()?,
ChangelogProducer::None | ChangelogProducer::Input
ChangelogProducer::None | ChangelogProducer::Input | ChangelogProducer::Lookup
)
|| (options.merge_engine()? == MergeEngine::FirstRow
&& options.try_changelog_producer()? != ChangelogProducer::None))
&& !matches!(
options.try_changelog_producer()?,
ChangelogProducer::None | ChangelogProducer::Lookup
)))
{
return Err(Error::Unsupported {
message: "PK compaction requires fixed or dynamic buckets and changelog-producer=none or input; first-row requires changelog-producer=none".into(),
message: "PK compaction requires fixed or dynamic buckets and changelog-producer=none, input or lookup; first-row requires changelog-producer=none or lookup".into(),
});
}
if options.try_changelog_producer()? == ChangelogProducer::Lookup {
primary_key::lookup::validate_options(properties)?;
}
options.validate_data_file_path_directory()
}

Expand Down
Loading
Loading