[
https://issues.apache.org/jira/browse/DRILL-3312?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14601815#comment-14601815
]
Jason Altekruse commented on DRILL-3312:
----------------------------------------
Do we ever expect users of this method to do anything with a null return
besides using it to trigger an out of memory exception? Is there any reason why
we want have the allocator do the null check and just throw an error rather
than ever returning null?
> PageReader.allocatePageData() calls BufferAllocator.buffer(int) but doesn't
> check if the result is null
> -------------------------------------------------------------------------------------------------------
>
> Key: DRILL-3312
> URL: https://issues.apache.org/jira/browse/DRILL-3312
> Project: Apache Drill
> Issue Type: Bug
> Components: Storage - Parquet
> Reporter: Deneche A. Hakim
> Assignee: Hanifi Gunes
> Fix For: 1.2.0
>
> Attachments: DRILL-3312.1.patch.txt
>
>
> Trying to reproduce DRILL-3241, Drill returns the following error:
> {noformat}
> SYSTEM ERROR: org.apache.drill.common.exceptions.DrillRuntimeException: Error
> in parquet record reader.
> Message:
> Hadoop path:
> /Users/hakim/MapR/data/tpcds100/parquet/store_sales/1_10_1.parquet
> Total records read: 0
> Mock records read: 0
> Records to read: 21845
> Row group index: 0
> Records in row group: 3348801
> Parquet Metadata: ParquetMetaData{FileMetaData{schema: message root {
> ...omitted text
> }
> , metadata: {}}, blocks: [...omitted text]}
> Fragment 3:0
> [Error Id: 0f9ba088-3091-4166-b46f-53ab60d364f8 on
> 10.250.50.54:31010]{noformat}
> The logs contain more information about the cause:
> {noformat}
> Caused by: java.lang.NullPointerException
> at
> org.apache.drill.exec.store.parquet.DirectCodecFactory$FullDirectDecompressor.decompress(DirectCodecFactory.java:238)
> ~[drill-java-exec-1.1.0-SNAPSHOT-rebuffed.jar:1.1.0-SNAPSHOT]
> at
> org.apache.drill.exec.store.parquet.columnreaders.PageReader.next(PageReader.java:230)
> ~[drill-java-exec-1.1.0-SNAPSHOT-rebuffed.jar:1.1.0-SNAPSHOT]
> at
> org.apache.drill.exec.store.parquet.columnreaders.NullableColumnReader.processPages(NullableColumnReader.java:76)
> ~[drill-java-exec-1.1.0-SNAPSHOT-rebuffed.jar:1.1.0-SNAPSHOT]
> at
> org.apache.drill.exec.store.parquet.columnreaders.ParquetRecordReader.readAllFixedFields(ParquetRecordReader.java:380)
> ~[drill-java-exec-1.1.0-SNAPSHOT-rebuffed.jar:1.1.0-SNAPSHOT]
> at
> org.apache.drill.exec.store.parquet.columnreaders.ParquetRecordReader.next(ParquetRecordReader.java:423)
> ~[drill-java-exec-1.1.0-SNAPSHOT-rebuffed.jar:1.1.0-SNAPSHOT]
> ...
> {noformat}
> Just before that line, PageReader.next() calls the following method:
> {noformat}
> private void allocatePageData(int size) {
> Preconditions.checkArgument(pageData == null);
> pageData =
> parentColumnReader.parentReader.getOperatorContext().getAllocator().buffer(size);
> }
> {noformat}
> this method should check the output of BufferAllocator.buffer(int) to make
> sure the buffer was successfully allocated.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
