[ 
https://issues.apache.org/jira/browse/HIVE-14621?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Sergey Shelukhin updated HIVE-14621:
------------------------------------
    Attachment: HIVE-14621.01.patch

Impl is still needed for other interfaces that it implements as we pass it on 
to other object. Renamed the call to startThreads for clarity... do you think 
we should start threads in ctor?

As for tests, we don't have any for this mode now. IO is initialized during 
driver init, so we'd need a separate CliDriver.

Btw, I was going to combine the interfaces for buffermanager and cache, since 
in both cases one object is used for both, but I couldn't come up with a name 
better than "CacheAndBufferManager", so I didn't do that.

> LLAP: memory.mode = none has NPE
> --------------------------------
>
>                 Key: HIVE-14621
>                 URL: https://issues.apache.org/jira/browse/HIVE-14621
>             Project: Hive
>          Issue Type: Bug
>            Reporter: Sergey Shelukhin
>            Assignee: Sergey Shelukhin
>         Attachments: HIVE-14621.01.patch, HIVE-14621.patch
>
>
> When IO elevator is enabled, but cache and allocator are both disabled, NPEs 
> happen. It's not really a recommended mode, but it's the only way to disable 
> cache, so we probably need to fix it. I am also going to nuke the 
> intermediate mode (allocator w/no cache) meanwhile cause it's pointless and 
> just creates a zoo of configurations.
> {noformat}
> Caused by: java.lang.NullPointerException
> at 
> org.apache.hadoop.hive.llap.cache.LlapDataBuffer.getByteBufferDup(LlapDataBuffer.java:59)
> at 
> org.apache.hadoop.hive.ql.io.orc.encoded.StreamUtils.createDiskRangeInfo(StreamUtils.java:63)
> at 
> org.apache.hadoop.hive.ql.io.orc.encoded.StreamUtils.createSettableUncompressedStream(StreamUtils.java:48)
> at 
> org.apache.hadoop.hive.ql.io.orc.encoded.EncodedTreeReaderFactory$LongStreamReader$StreamReaderBuilder.build(EncodedTreeReaderFactory.java:514)
> at 
> org.apache.hadoop.hive.ql.io.orc.encoded.EncodedTreeReaderFactory.createEncodedTreeReader(EncodedTreeReaderFactory.java:1737)
> at 
> org.apache.hadoop.hive.llap.io.decode.OrcEncodedDataConsumer.decodeBatch(OrcEncodedDataConsumer.java:162)
> at 
> org.apache.hadoop.hive.llap.io.decode.OrcEncodedDataConsumer.decodeBatch(OrcEncodedDataConsumer.java:55)
> at 
> org.apache.hadoop.hive.llap.io.decode.EncodedDataConsumer.consumeData(EncodedDataConsumer.java:76)
> at 
> org.apache.hadoop.hive.llap.io.decode.EncodedDataConsumer.consumeData(EncodedDataConsumer.java:30)
> at 
> org.apache.hadoop.hive.ql.io.orc.encoded.EncodedReaderImpl.readEncodedColumns(EncodedReaderImpl.java:408)
> at 
> org.apache.hadoop.hive.llap.io.encoded.OrcEncodedDataReader.performDataRead(OrcEncodedDataReader.java:424)
> at 
> org.apache.hadoop.hive.llap.io.encoded.OrcEncodedDataReader$4.run(OrcEncodedDataReader.java:227)
> at 
> org.apache.hadoop.hive.llap.io.encoded.OrcEncodedDataReader$4.run(OrcEncodedDataReader.java:224)
> at java.security.AccessController.doPrivileged(Native Method)
> at javax.security.auth.Subject.doAs(Subject.java:415)
> at 
> org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1724)
> at 
> org.apache.hadoop.hive.llap.io.encoded.OrcEncodedDataReader.callInternal(OrcEncodedDataReader.java:224)
> at 
> org.apache.hadoop.hive.llap.io.encoded.OrcEncodedDataReader.callInternal(OrcEncodedDataReader.java:93)
> ... 6 more
> {noformat}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to