Affects Version/s: 1.11.3, 1.12.0, 1.13.0
Fix Version/s: None
Flink version: 1.10
StateBackend : FsStateBackend
code: Flink SQL count(distinct userId)
uv: 10 million
State size: 200M
TM total memory: 16G
Heap StateBackend needs to serialize each Java Object into the file system during snapshot. RocksDB StateBackend's RocksFullSnapshotStrategy needs to read kvs from RocksDB and write them to the file system in the snapshot.
The above two cases involve a lot of small io, not large io, frequent small io is not friendly to disk. Therefore, the buffer is used in the checkpoint snapshot writing process of the file system. For details, refer to the buffer of FsCheckpointStreamFactory.FsCheckpointStateOutputStream.
There will be many small IOs in the restore process, but restore does not have a buffer. So I added a buffer and tested it based on Flink job.
It takes 33.1s to restore without read buffer, and 12.8s to restore with read buffer.
Use FSDataBufferedInputStream to wrap fsDataInputStream in HeapRestoreOperation#restore，code：