Github user QiangCai commented on a diff in the pull request:
https://github.com/apache/incubator-carbondata/pull/203#discussion_r82334120
--- Diff:
processing/src/main/java/org/apache/carbondata/processing/csvreaderstep/CsvInput.java
---
@@ -343,41 +349,57 @@ public boolean processRow(StepMetaInterface smi,
StepDataInterface sdi) throws K
} catch (NumberFormatException exc) {
numberOfNodes = NUM_CORES_DEFAULT_VAL;
}
+ if ( rddIteratorKey == null ) {
+ BlockDetails[] blocksInfo =
GraphGenerator.blockInfo.get(meta.getBlocksID());
+ if (blocksInfo.length == 0) {
+ //if isDirectLoad = true, and partition number > file num
+ //then blocksInfo will get empty in some partition processing, so
just return
+ setOutputDone();
+ return false;
+ }
- BlockDetails[] blocksInfo =
GraphGenerator.blockInfo.get(meta.getBlocksID());
- if (blocksInfo.length == 0) {
- //if isDirectLoad = true, and partition number > file num
- //then blocksInfo will get empty in some partition processing, so
just return
- setOutputDone();
- return false;
- }
-
- if (numberOfNodes > blocksInfo.length) {
- numberOfNodes = blocksInfo.length;
- }
+ if (numberOfNodes > blocksInfo.length) {
+ numberOfNodes = blocksInfo.length;
+ }
- //new the empty lists
- for (int pos = 0; pos < numberOfNodes; pos++) {
- threadBlockList.add(new ArrayList<BlockDetails>());
- }
+ //new the empty lists
+ for (int pos = 0; pos < numberOfNodes; pos++) {
+ threadBlockList.add(new ArrayList<BlockDetails>());
+ }
- //block balance to every thread
- for (int pos = 0; pos < blocksInfo.length; ) {
- for (int threadNum = 0; threadNum < numberOfNodes; threadNum++) {
- if (pos < blocksInfo.length) {
- threadBlockList.get(threadNum).add(blocksInfo[pos++]);
+ //block balance to every thread
+ for (int pos = 0; pos < blocksInfo.length; ) {
+ for (int threadNum = 0; threadNum < numberOfNodes; threadNum++) {
+ if (pos < blocksInfo.length) {
+ threadBlockList.get(threadNum).add(blocksInfo[pos++]);
+ }
}
}
+ LOGGER.info("*****************Started all csv reading***********");
+ startProcess(numberOfNodes);
+ LOGGER.info("*****************Completed all csv reading***********");
+
CarbonTimeStatisticsFactory.getLoadStatisticsInstance().recordCsvInputStepTime(
+ meta.getPartitionID(), System.currentTimeMillis());
+ } else {
+ scanRddIterator();
}
- LOGGER.info("*****************Started all csv reading***********");
- startProcess(numberOfNodes);
- LOGGER.info("*****************Completed all csv reading***********");
-
CarbonTimeStatisticsFactory.getLoadStatisticsInstance().recordCsvInputStepTime(
- meta.getPartitionID(), System.currentTimeMillis());
setOutputDone();
return false;
}
+ private void scanRddIterator() throws RuntimeException {
+ Iterator<String[]> iterator =
RddInputUtils.getAndRemove(rddIteratorKey);
+ if (iterator != null) {
+ try{
+ while(iterator.hasNext()){
+ putRow(data.outputRowMeta, iterator.next());
--- End diff --
I agree, we can merge kettle steps.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
