Skip to content

[SPARK-48586][SS] Remove lock acquisition in doMaintenance() by making a deep copy of file mappings in RocksDBFileManager in load() #46942

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

Closed
riyaverm-db wants to merge 18 commits into from
Closed

[SPARK-48586][SS] Remove lock acquisition in doMaintenance() by making a deep copy of file mappings in RocksDBFileManager in load() #46942

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
18 commits
Select commit Hold shift + click to select a range
b8bf6cc
Add deepCopy() to RocksDBFileManager
riyaverm-db Jun 3, 2024
d9cfc70
Add deepCopy test case
riyaverm-db Jun 3, 2024
1b64014
Fix Formatting
riyaverm-db Jun 4, 2024
28451ca
Revert pr changes
riyaverm-db Jun 5, 2024
a995091
Use deepCopy to avoid lock contention
riyaverm-db Jun 7, 2024
25b9e96
Modify deep copy test case, add comments
riyaverm-db Jun 12, 2024
d3cbf25
Add additional asserts to deepCopy test case
riyaverm-db Jun 12, 2024
0bf9acd
Add lock contention test case
riyaverm-db Jun 17, 2024
6069779
Add latestSnapshot race condition test case
riyaverm-db Jun 18, 2024
7381b17
Modify test case comments
riyaverm-db Jun 18, 2024
651ac74
Move lock contention test case
riyaverm-db Jun 18, 2024
d3354cf
Modify lock test case
riyaverm-db Jun 19, 2024
7b4ebd6
Remove latestSnapshot race condition and refactor RocksDBFileManager …
riyaverm-db Jun 24, 2024
c57f958
Purge oldSnapshots one at a time
riyaverm-db Jun 24, 2024
ddb97d6
Minor code style and naming changes
riyaverm-db Jun 25, 2024
5894a97
Add synchronization in commit to remove latestSnapshot race condition
riyaverm-db Jun 26, 2024
bcb2307
Remove Option in list type
riyaverm-db Jun 26, 2024
16c26fc
Remove extra line
riyaverm-db Jun 26, 2024
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
Prev Previous commit
Next Next commit
Minor code style and naming changes
  • Loading branch information
@riyaverm-db
riyaverm-db committed Jun 25, 2024
commit ddb97d66a4fd04f873dc5386b075be0ceb927cfd
4 changes: 2 additions & 2 deletions sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/state/RocksDB.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -78,7 +78,7 @@ class RocksDB(
checkpointDir: File,
Copy link
Contributor

@HeartSaVioR HeartSaVioR Jun 27, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: 2 more spaces (4 spaces for params in multi-lines)

version: Long,
numKeys: Long,
fileMappings: RocksDBFileMappings) {
capturedFileMappings: RocksDBFileMappings) {
def close(): Unit = {
silentDeleteRecursively(checkpointDir, s"Free up local checkpoint of snapshot $version")
}
Expand Down Expand Up @@ -596,7 +596,7 @@ class RocksDB(
// inside the uploadSnapshot() called below.
// If changelog checkpointing is enabled, snapshot will be uploaded asynchronously
// during state store maintenance.
oldSnapshots.addOne(latestSnapshot)
oldSnapshots += latestSnapshot
Copy link
Contributor

@sahnib sahnib Jun 26, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Do we need to check if latestSnapshot is None before we append it to the list?

Copy link
Contributor

@chaoqin-li1123 chaoqin-li1123 Jun 26, 2024

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is a List[Option[Snapshot]], so checking is not needed. But I feel changing it to a List[Snapshot] and doing the null check make the code much cleaner.

latestSnapshot = Some(
RocksDBSnapshot(checkpointDir,
newVersion,
Expand Down
2 changes: 1 addition & 1 deletion ...re/src/main/scala/org/apache/spark/sql/execution/streaming/state/RocksDBFileManager.scala
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -770,7 +770,7 @@ class RocksDBFileManager(

/**
* Track file mappings in RocksDB across local and remote directories
* @param versionToRocksDBFiles Mapping of version to DFS files
* @param versionToRocksDBFiles Mapping of RocksDB files used across versions for maintenance
* @param localFilesToDfsFiles Mapping of the exact Dfs file used to create a local SST file
* The reason localFilesToDfsFiles is a separate map because versionToRocksDBFiles can contain
* multiple similar SST files to a particular local file (for example 1.sst can map to 1-UUID1.sst
Expand Down