lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

Adding a config to control whether date partitioning can be assumed

Browse Source 
- false by default
 - CAUTION: If you have an existing tables without partition metadata, you need to set this to "true"
This commit is contained in:
Vinoth Chandar
2017-03-26 17:40:20 -07:00
committed by vinoth chandar
parent f9fd16069d
commit dce35ff0d7
7 changed files with 32 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
hoodie-client/src/main/java/com/uber/hoodie/HoodieWriteClient.java
View File

@@ -476,7 +476,7 @@ public class HoodieWriteClient<T extends HoodieRecordPayload> implements Seriali
+ lastCommitRetained);
Map<String, List<String>> latestFilesMap = jsc.parallelize(
FSUtils.getAllPartitionPaths(fs, table.getMetaClient().getBasePath()))
FSUtils.getAllPartitionPaths(fs, table.getMetaClient().getBasePath(), config.shouldAssumeDatePartitioning()))
.mapToPair((PairFunction<String, String, List<String>>) partitionPath -> {
// Scan all partitions files with this commit time
logger.info("Collecting latest files in partition path " + partitionPath);
@@ -650,7 +650,7 @@ public class HoodieWriteClient<T extends HoodieRecordPayload> implements Seriali
logger.info("Clean out all parquet files generated for commits: " + commits);
final LongAccumulator numFilesDeletedCounter = jsc.sc().longAccumulator();
List<HoodieRollbackStat> stats = jsc.parallelize(
FSUtils.getAllPartitionPaths(fs, table.getMetaClient().getBasePath()))
FSUtils.getAllPartitionPaths(fs, table.getMetaClient().getBasePath(), config.shouldAssumeDatePartitioning()))
.map((Function<String, HoodieRollbackStat>) partitionPath -> {
// Scan all partitions files with this commit time
logger.info("Cleaning path " + partitionPath);
@@ -739,7 +739,7 @@ public class HoodieWriteClient<T extends HoodieRecordPayload> implements Seriali
.getHoodieTable(new HoodieTableMetaClient(fs, config.getBasePath(), true), config);
List<String> partitionsToClean =
FSUtils.getAllPartitionPaths(fs, table.getMetaClient().getBasePath());
FSUtils.getAllPartitionPaths(fs, table.getMetaClient().getBasePath(), config.shouldAssumeDatePartitioning());
// shuffle to distribute cleaning work across partitions evenly
Collections.shuffle(partitionsToClean);
logger.info("Partitions to clean up : " + partitionsToClean + ", with policy " + config