1996fanrui opened a new pull request #13885:
URL: https://github.com/apache/flink/pull/13885
## What is the purpose of the change
Read checkpoint stream with buffer to speedup restore.
There will be many small IOs in the restore process, frequent small io is
not friendly to disk. So add buffer to improve IO performance.
## Brief change log
- add config `state.backend.fs.read-buffer-size`
- Design FSDataBufferedInputStream, FSDataInputStream with buffer
- Use FSDataBufferedInputStream to wrap fsDataInputStream in
HeapRestoreOperation#restore.
tips:
The design of FSDataBufferedInputStream refers to
FsCheckpointStateOutputStream and Java's BufferedInputStream.
Compared to Java's BufferedInputStream, the seek and getPos interfaces are
mainly added.
## Verifying this change
This change added tests and can be verified as follows:
- Added order read for FSDataBufferedInputStream
- Added seek and random read for FSDataBufferedInputStream
## Does this pull request potentially affect one of the following parts:
- Dependencies (does it add or upgrade a dependency): no
- The public API, i.e., is any changed class annotated with
`@Public(Evolving)`: no
- The serializers: no
- The runtime per-record code paths (performance sensitive): no
- Anything that affects deployment or recovery: JobManager (and its
components), Checkpointing, Kubernetes/Yarn/Mesos, ZooKeeper: yes
- The S3 file system connector: no
## Documentation
- Does this pull request introduce a new feature? yes
- If yes, how is the feature documented? JavaDocs
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
