- Fixes#246
- Bump up default parallelism to 1500, to handle large upserts
- Add docs on s3 confuration & tuning tips with tested spark knobs
- Fix bug to not duplicate hoodie metadata fields when input dataframe is another hoodie dataset
- Improve speed of ROTablePathFilter by removing directory check
- Move to spark-avro 4.0 to handle issue with nested fields with same name
- Keep AvroConversionUtils in sync with spark-avro 4.0
The code-style rules follow google style with some changes:
1. Increase line length from 100 to 120
2. Disable JavaDoc related checkstyles as this needs more manual work.
Both source and test code are checked for code-style
- Introduced concept of converters to be able to serde generic datatype for SpillableMap
- Fixed/Added configs to Hoodie Configs
- Changed HoodieMergeHandle to start using SpillableMap
- HoodieLogFormat V2 has support for LogFormat evolution through versioning
- LogVersion is associated with a LogBlock not a LogFile
- Based on a version for a LogBlock, approporiate code path is executed
- Implemented LazyReading of Hoodie Log Blocks with Memory / IO tradeoff
- Implemented Reverse pointer to be able to traverse the log in reverse
- Introduce new MAGIC for backwards compatibility with logs without versions
Use FastDateFormat for thread safety, this is to fix an exception when a
job is used to ingest multiple tables. An example exception:
```
Caused by: java.lang.NumberFormatException: multiple points
at sun.misc.FloatingDecimal.readJavaFormatString(FloatingDecimal.java:1890)
at sun.misc.FloatingDecimal.parseDouble(FloatingDecimal.java:110)
at java.lang.Double.parseDouble(Double.java:538)
at java.text.DigitList.getDouble(DigitList.java:169)
at java.text.DecimalFormat.parse(DecimalFormat.java:2056)
at java.text.SimpleDateFormat.subParse(SimpleDateFormat.java:1867)
at java.text.SimpleDateFormat.parse(SimpleDateFormat.java:1514)
at java.text.DateFormat.parse(DateFormat.java:364)
at com.uber.hoodie.HoodieWriteClient.commit(HoodieWriteClient.java:442)
```
- Fallback to old way of rollback by listing all partitions
- Added null check to ensure only partitions which are to be rolledback are considered
- Added location (committime) to workload stat
- Added checks in CompactedScanner to guard against task retries
- Introduce new logic for rollback (bounded by instant_time and target_instant time)
- Reversed logfiles order
- When append() is not supported, rollover to new file always (instead of failing)
- Provide way to configure archive log folder (avoids small files inside .hoodie)
- Datasets written via Spark datasource archive to .hoodie/archived
- HoodieClientExample will now retain only 2,3 commits to exercise archival path during dev cycles
- Few tweaks to code structure around CommitArchiveLog
- Reviving PR 191, to make FileSystem creation off actual path
- Streamline all filesystem access to HoodieTableMetaClient
- Hadoop Conf from Spark Context serialized & passed to executor code too
- Pick up env vars prefixed with HOODIE_ENV_ into Configuration object
- Cleanup usage of FSUtils.getFS, piggybacking off HoodieTableMetaClient.getFS
- Adding s3a to supported schemes & support escaping "." in env vars
- Tests use HoodieTestUtils.getDefaultHadoopConf
- Write with COW/MOR paths work fully
- Read with RO view works on both storages*
- Incremental view supported on COW
- Refactored out HoodieReadClient methods, to just contain key based access
- HoodieDataSourceHelpers class can be now used to construct inputs to datasource
- Tests in hoodie-client using new helpers and mechanisms
- Basic tests around save modes & insert/upserts (more to follow)
- Bumped up scala to 2.11, since 2.10 is deprecated & complains with scalatest
- Updated documentation to describe usage
- New sample app written using the DataSource API
- keys compared lexicographically using String::compareTo
- Range metadata additionally written into parquet file footers
- Trim fat & few optimizations to speed up indexing
- Add param to control whether input shall be cached, to speed up lookup
- Add param to turn on/off range pruning
- Auto compute of parallelism now simply factors in amount of comparisons done
- More accurate parallelism computation when range pruning is on
- tests added & hardened, docs updated
