You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@flink.apache.org by GitBox <gi...@apache.org> on 2018/10/18 09:56:25 UTC

[GitHub] tzulitai opened a new pull request #6875: [FLINK-9808] [state backends] Migrate state when necessary in state backends

tzulitai opened a new pull request #6875: [FLINK-9808] [state backends] Migrate state when necessary in state backends
URL: https://github.com/apache/flink/pull/6875
 
 
   ## What is the purpose of the change
   
   This PR adds the procedure of migrating state (i.e., reading state bytes with the restored prior serializer, and then re-writing the deserialized object with the new serializer).
   
   This should only ever occur in the RocksDB state backend, since the restore / snapshotting of heap-based backends are already a migration process by nature.
   
   The requirement to perform a state migration is decided by the `CompatibilityResult` returned from compatibility checks against the new serializer using its prior serializer's snapshot.
   
   ## Brief change log
   
   - Add `migrateSerializedValue` to `AbstractRocksDBStateValue`
   - Use the new method to iterate through RocksDB bytes when state migration is necessary
   - Remove checks that are no longer relevant in the heap backends, since it is inherently already migrating state bytes as part of its restore / snapshot process
   - Add a `StateMigrationTestBase` that verifies serializers are being used as expected for a restore procedure
   
   ## Verifying this change
   
   The main new test coverage that covers this is in the `StateMigrationTestBase` class.
   
   ## Does this pull request potentially affect one of the following parts:
   
     - Dependencies (does it add or upgrade a dependency): (yes / **no**)
     - The public API, i.e., is any changed class annotated with `@Public(Evolving)`: (yes / **no**)
     - The serializers: (yes / **no** / don't know)
     - The runtime per-record code paths (performance sensitive): (**yes** / no / don't know)
     - Anything that affects deployment or recovery: JobManager (and its components), Checkpointing, Yarn/Mesos, ZooKeeper: (**yes** / no / don't know)
     - The S3 file system connector: (yes / **no** / don't know)
   
   ## Documentation
   
     - Does this pull request introduce a new feature? (**yes** / no)
     - If yes, how is the feature documented? (**not yet documented**)

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services