Addresses leaks, perf degradation observed during testing. These were regressions from the original rfc-15 PoC implementation.
* Pass a single instance of HoodieTableMetadata everywhere
* Fix tests and add config for enabling metrics
- Removed special casing of assumeDatePartitioning inside FSUtils#getAllPartitionPaths()
- Consequently, IOException is never thrown and many files had to be adjusted
- More diligent handling of open file handles in metadata table
- Added config for controlling reuse of connections
- Added config for turning off fallback to listing, so we can see tests fail
- Changed all ipf listing code to cache/amortize the open/close for better performance
- Timelineserver also reuses connections, for better performance
- Without timelineserver, when metadata table is opened from executors, reuse is not allowed
- HoodieMetadataConfig passed into HoodieTableMetadata#create as argument.
- Fix TestHoodieBackedTableMetadata#testSync
* [HUDI-1479] Use HoodieEngineContext to parallelize fetching of partition paths
* Adding testClass for FileSystemBackedTableMetadata
Co-authored-by: Nishith Agarwal <nagarwal@uber.com>
* [HUDI-1481] add structured streaming and delta streamer clustering unit test
* [HUDI-1399] support a independent clustering spark job to asynchronously clustering
* [HUDI-1399] support a independent clustering spark job to asynchronously clustering
* [HUDI-1498] Read clustering plan from requested file for inflight instant (#2389)
* [HUDI-1399] support a independent clustering spark job with schedule generate instant time
Co-authored-by: satishkotha <satishkotha@uber.com>
* Added ability to pass in `properties` to payload methods, so they can perform table/record specific merges
* Added default methods so existing payload classes are backwards compatible.
* Adding DefaultHoodiePayload to honor ordering while merging two records
* Fixing default payload based on feedback
* Fix flaky MOR unit test
* Update Spark APIs to make it be compatible with both spark2 & spark3
* Refactor bulk insert v2 part to make Hudi be able to compile with Spark3
* Add spark3 profile to handle fasterxml & spark version
* Create hudi-spark-common module & refactor hudi-spark related modules
Co-authored-by: Wenning Ding <wenningd@amazon.com>
The current DFSPathSelector only ignore prefix(_, .) at the file level while files under subdirectories
e.g. (.checkpoint/*) are still considered which result in bad-format exception during reading.
1. Use the DAG Node's label from the yaml as its name instead of UUID names which are not descriptive when debugging issues from logs.
2. Fix CleanNode constructor which is not correctly implemented
3. When generating upsets, allows more granualar control over the number of inserts and upserts - zero or more inserts and upserts can be specified instead of always requiring both inserts and upserts.
4. Fixed generation of records of specific size
- The current code was using a class variable "shouldAddMore" which was reset to false after the first record generation causing subsequent records to be of minimum size.
- In this change, we pre-calculate the extra size of the complex fields. When generating records, for complex fields we read the field size from this map.
5. Refresh the timeline of the DeltaSync service before calling readFromSource. This ensures that only the newest generated data is read and data generated in the older Dag Nodes is ignored (as their AVRO files will have an older timestamp).
6. Making --workload-generator-classname an optional parameter as most probably the default will be used
- This change breaks `hudi-client` into `hudi-client-common` and `hudi-spark-client` modules
- Simple usages of Spark using jsc.parallelize() has been redone using EngineContext#map, EngineContext#flatMap etc
- Code changes in the PR, break classes into `BaseXYZ` parent classes with no spark dependencies living in `hudi-client-common`
- Classes on `hudi-spark-client` are named `SparkXYZ` extending the parent classes with all the Spark dependencies
- To simplify/cleanup, HoodieIndex#fetchRecordLocation has been removed and its usages in tests replaced with alternatives
Co-authored-by: Vinoth Chandar <vinoth@apache.org>
* [HUDI-1181] Fix decimal type display issue for record key field
* Remove getNestedFieldVal method from DataSourceUtils
* resolve comments
Co-authored-by: Wenning Ding <wenningd@amazon.com>
For Delete API, "hoodie.delete.shuffle.parallelism" isn't used as opposed to "hoodie.upsert.shuffle.parallelism" is used for upsert, this creates the performance difference between delete by upsert API with "EmptyHoodieRecordPayload" and delete API for certain cases.
This patch makes the following fixes in this regard.
- Let deduplicateKeys method use "hoodie.delete.shuffle.parallelism"
- Repartition inputRDD as "hoodie.delete.shuffle.parallelism" in case "hoodie.combine.before.delete=false"
- Generalize the hive-sync module for syncing to multiple metastores
- Added new options for datasource
- Added new command line for delta streamer
Co-authored-by: Vinoth Chandar <vinoth@apache.org>
- [HUDI-418] Bootstrap Index Implementation using HFile with unit-test
- [HUDI-421] FileSystem View Changes to support Bootstrap with unit-tests
- [HUDI-424] Implement Query Side Integration for querying tables containing bootstrap file slices
- [HUDI-423] Implement upsert functionality for handling updates to these bootstrap file slices
- [HUDI-421] Bootstrap Write Client with tests
- [HUDI-425] Added HoodieDeltaStreamer support
- [HUDI-899] Add a knob to change partition-path style while performing metadata bootstrap
- [HUDI-900] Metadata Bootstrap Key Generator needs to handle complex keys correctly
- [HUDI-424] Simplify Record reader implementation
- [HUDI-423] Implement upsert functionality for handling updates to these bootstrap file slices
- [HUDI-420] Hoodie Demo working with hive and sparkSQL. Also, Hoodie CLI working with bootstrap tables
Co-authored-by: Mehrotra <uditme@amazon.com>
Co-authored-by: Vinoth Chandar <vinoth@apache.org>
Co-authored-by: Balaji Varadarajan <varadarb@uber.com>
* [HUDI-839] Introducing rollback strategy using marker files
- Adds a new mechanism for rollbacks where it's based on the marker files generated during the write
- Consequently, marker file/dir deletion now happens post commit, instead of during finalize
- Marker files are also generated for AppendHandle, making it consistent throughout the write path
- Until upgrade-downgrade mechanism can upgrade non-marker based inflight writes to marker based, this should only be turned on for new datasets.
- Added marker dir deletion after successful commit/rollback, individual files are not deleted during finalize
- Fail safe for deleting marker directories, now during timeline archival process
- Added check to ensure completed instants are not rolled back using marker based strategy. This will be incorrect
- Reworked tests to rollback inflight instants, instead of completed instants whenever necessary
- Added an unit test for MarkerBasedRollbackStrategy
Co-authored-by: Vinoth Chandar <vinoth@apache.org>
