lanyuanxiaoyao/hudi
1
0
Fork 0
Code Activity

1. Small file size handling for inserts into log files. In summary, the total size of the log file is compared with the parquet max file size and if there is scope to add inserts the add it.

Browse Source
This commit is contained in:
Nishith Agarwal
2018-06-11 20:27:56 -07:00
committed by vinoth chandar
parent 324de298bc
commit 459e523d9e
21 changed files with 824 additions and 123 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
hoodie-hive/src/main/java/com/uber/hoodie/hive/HoodieHiveClient.java
View File

@@ -307,7 +307,7 @@ public class HoodieHiveClient {
HoodieInstant lastCommit = activeTimeline.lastInstant().orElseThrow(
() -> new InvalidDatasetException(syncConfig.basePath));
HoodieCommitMetadata commitMetadata = HoodieCommitMetadata.fromBytes(
activeTimeline.getInstantDetails(lastCommit).get());
activeTimeline.getInstantDetails(lastCommit).get(), HoodieCommitMetadata.class);
String filePath = commitMetadata.getFileIdAndFullPaths(metaClient.getBasePath()).values()
.stream().findAny().orElseThrow(() -> new IllegalArgumentException(
"Could not find any data file written for commit " + lastCommit
@@ -339,7 +339,7 @@ public class HoodieHiveClient {
HoodieInstant lastDeltaInstant = lastDeltaCommit.get();
// read from the log file wrote
commitMetadata = HoodieCommitMetadata.fromBytes(
activeTimeline.getInstantDetails(lastDeltaInstant).get());
activeTimeline.getInstantDetails(lastDeltaInstant).get(), HoodieCommitMetadata.class);
filePath = commitMetadata.getFileIdAndFullPaths(metaClient.getBasePath()).values()
.stream().filter(s -> s.contains(HoodieLogFile.DELTA_EXTENSION))
.findAny().orElseThrow(() -> new IllegalArgumentException(
@@ -372,7 +372,7 @@ public class HoodieHiveClient {
// Read from the compacted file wrote
HoodieCommitMetadata compactionMetadata = HoodieCommitMetadata.fromBytes(
activeTimeline.getInstantDetails(lastCompactionCommit).get());
activeTimeline.getInstantDetails(lastCompactionCommit).get(), HoodieCommitMetadata.class);
String filePath = compactionMetadata.getFileIdAndFullPaths(metaClient.getBasePath()).values()
.stream().findAny().orElseThrow(() -> new IllegalArgumentException(
"Could not find any data file written for compaction " + lastCompactionCommit
@@ -539,7 +539,7 @@ public class HoodieHiveClient {
Integer.MAX_VALUE);
return timelineToSync.getInstants().map(s -> {
try {
return HoodieCommitMetadata.fromBytes(activeTimeline.getInstantDetails(s).get());
return HoodieCommitMetadata.fromBytes(activeTimeline.getInstantDetails(s).get(), HoodieCommitMetadata.class);
} catch (IOException e) {
throw new HoodieIOException(
"Failed to get partitions written since " + lastCommitTimeSynced, e);