Skip to content
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.

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

fix(storage): fix start table key panic when there are only range-tombstones #8189

Merged
merged 3 commits into from
Feb 27, 2023
Merged
Changes from all commits
Commits
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
60 changes: 35 additions & 25 deletions src/storage/src/hummock/shared_buffer/shared_buffer_batch.rs
Original file line number Diff line number Diff line change
Expand Up @@ -44,6 +44,7 @@ pub(crate) struct SharedBufferBatchInner {
payload: Vec<SharedBufferItem>,
range_tombstone_list: Vec<DeleteRangeTombstone>,
largest_table_key: Vec<u8>,
smallest_table_key: Vec<u8>,
size: usize,
_tracker: Option<MemoryTracker>,
batch_id: SharedBufferBatchId,
Expand All @@ -57,6 +58,8 @@ impl SharedBufferBatchInner {
_tracker: Option<MemoryTracker>,
) -> Self {
let mut largest_table_key = vec![];
let mut smallest_table_key = vec![];
let mut smallest_empty = true;
if !range_tombstone_list.is_empty() {
range_tombstone_list.sort();
let mut range_tombstones: Vec<DeleteRangeTombstone> = vec![];
Expand All @@ -68,6 +71,11 @@ impl SharedBufferBatchInner {
largest_table_key.clear();
largest_table_key.extend_from_slice(&tombstone.end_user_key.table_key.0);
}
if smallest_empty || smallest_table_key.gt(&tombstone.start_user_key.table_key.0) {
smallest_table_key.clear();
smallest_table_key.extend_from_slice(&tombstone.start_user_key.table_key.0);
smallest_empty = false;
}
if let Some(last) = range_tombstones.last_mut() {
if last.end_user_key.gt(&tombstone.start_user_key) {
if last.end_user_key.lt(&tombstone.end_user_key) {
Expand All @@ -87,11 +95,18 @@ impl SharedBufferBatchInner {
largest_table_key.extend_from_slice(item.0.as_ref());
}
}
if let Some(item) = payload.first() {
if smallest_empty || item.0.lt(&smallest_table_key) {
smallest_table_key.clear();
smallest_table_key.extend_from_slice(item.0.as_ref());
}
}
SharedBufferBatchInner {
payload,
range_tombstone_list,
size,
largest_table_key,
smallest_table_key,
_tracker,
batch_id: SHARED_BUFFER_BATCH_ID_GENERATOR.fetch_add(1, Relaxed),
}
Expand Down Expand Up @@ -242,39 +257,20 @@ impl SharedBufferBatch {
&self.inner
}

#[inline(always)]
pub fn start_table_key(&self) -> TableKey<&[u8]> {
TableKey(&self.inner.first().unwrap().0)
TableKey(&self.inner.smallest_table_key)
}

#[inline(always)]
pub fn end_table_key(&self) -> TableKey<&[u8]> {
TableKey(&self.inner.last().unwrap().0)
TableKey(&self.inner.largest_table_key)
}

/// return inclusive left endpoint, which means that all data in this batch should be larger or
/// equal than this key.
pub fn start_user_key(&self) -> UserKey<&[u8]> {
if self.has_range_tombstone()
&& (self.inner.is_empty()
|| self
.inner
.range_tombstone_list
.first()
.unwrap()
.start_user_key
.table_key
.0
.as_slice()
.le(&self.inner.first().unwrap().0))
{
self.inner
.range_tombstone_list
.first()
.unwrap()
.start_user_key
.as_ref()
} else {
UserKey::new(self.table_id, self.start_table_key())
}
UserKey::new(self.table_id, self.start_table_key())
}

#[inline(always)]
Expand All @@ -285,7 +281,7 @@ impl SharedBufferBatch {
/// return inclusive right endpoint, which means that all data in this batch should be smaller
/// or equal than this key.
pub fn end_user_key(&self) -> UserKey<&[u8]> {
UserKey::new(self.table_id, TableKey(&self.inner.largest_table_key))
UserKey::new(self.table_id, self.end_table_key())
}

pub fn epoch(&self) -> u64 {
Expand Down Expand Up @@ -620,6 +616,20 @@ mod tests {
}
output.reverse();
assert_eq!(output, shared_buffer_items);

let batch = SharedBufferBatch::build_shared_buffer_batch(
epoch,
vec![],
1,
vec![
(Bytes::from("a"), Bytes::from("c")),
(Bytes::from("b"), Bytes::from("d")),
],
TableId::new(0),
None,
);
assert_eq!(batch.start_table_key().as_ref(), "a".as_bytes());
assert_eq!(batch.end_table_key().as_ref(), "d".as_bytes());
}

#[tokio::test]
Expand Down