You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@drill.apache.org by "Kunal Khatua (JIRA)" <ji...@apache.org> on 2018/04/26 22:07:00 UTC

[jira] [Updated] (DRILL-6235) Flatten query leads to out of memory in RPC layer.

     [ https://issues.apache.org/jira/browse/DRILL-6235?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Kunal Khatua updated DRILL-6235:
--------------------------------
    Affects Version/s: 1.13.0

> Flatten query leads to out of memory in RPC layer.
> --------------------------------------------------
>
>                 Key: DRILL-6235
>                 URL: https://issues.apache.org/jira/browse/DRILL-6235
>             Project: Apache Drill
>          Issue Type: Bug
>          Components: Execution - Flow
>    Affects Versions: 1.12.0, 1.13.0
>            Reporter: Khurram Faraaz
>            Assignee: Padma Penumarthy
>            Priority: Critical
>         Attachments: 25593391-512d-23ab-7c84-3651006931e2.sys.drill, drillbit_snippet.log
>
>
> Flatten query leads to out of memory in RPC layer. Query profile is attached here.
> Total number of JSON files = 4095
>  Each JSON file has nine rows
>  And each row in the JSON has an array with 1024 integer values, and there are other string values outside of the array.
>  Two major fragments and eighty eight minor fragments were created
> On a 4 node CentOS cluster
>  number of CPU cores
>  [root@qa102-45 ~]# grep -c ^processor /proc/cpuinfo
>  32
> Details of memory
> {noformat}
> 0: jdbc:drill:schema=dfs.tmp> select * from sys.memory;
> +------------------+------------+---------------+-------------+-----------------+---------------------+-------------+
> | hostname | user_port | heap_current | heap_max | direct_current | jvm_direct_current | direct_max |
> +------------------+------------+---------------+-------------+-----------------+---------------------+-------------+
> | qa102-45.qa.lab | 31010 | 1130364912 | 4294967296 | 0 | 170528 | 8589934592 |
> | qa102-47.qa.lab | 31010 | 171823104 | 4294967296 | 0 | 21912 | 8589934592 |
> | qa102-48.qa.lab | 31010 | 201326576 | 4294967296 | 0 | 21912 | 8589934592 |
> | qa102-46.qa.lab | 31010 | 214780896 | 4294967296 | 0 | 21912 | 8589934592 |
> +------------------+------------+---------------+-------------+-----------------+---------------------+-------------+
> 4 rows selected (0.166 seconds)
> {noformat}
> Steps to repro the failure.
> Reset all options and set slice_target=1 and run the below SQL query.
>  alter system reset all;
>  alter system set `planner.slice_target`=1;
> {noformat}
> SELECT * , FLATTEN(arr) FROM many_json_files
> ...
> Error: RESOURCE ERROR: One or more nodes ran out of memory while executing the query.
> Failure allocating buffer.
> Fragment 1:38
> [Error Id: cf4fd273-d8a2-45e8-8d72-15c738e53b0f on qa102-45.qa.lab:31010] (state=,code=0)
> {noformat}
> Stack trace from drillbit.log fir above failing query.
> {noformat}
> 2018-03-12 11:52:33,849 [25593391-512d-23ab-7c84-3651006931e2:frag:0:0] INFO o.a.d.e.w.fragment.FragmentExecutor - 25593391-512d-23ab-7c84-3651006931e2:0:0: State change requested AWAITING_ALLOCATION --> RUNNING
> 2018-03-12 11:52:33,849 [25593391-512d-23ab-7c84-3651006931e2:frag:0:0] INFO o.a.d.e.w.f.FragmentStatusReporter - 25593391-512d-23ab-7c84-3651006931e2:0:0: State to report: RUNNING
> 2018-03-12 11:52:33,854 [25593391-512d-23ab-7c84-3651006931e2:frag:0:0] INFO o.a.d.e.c.ClassCompilerSelector - Java compiler policy: DEFAULT, Debug option: true
> 2018-03-12 11:52:35,929 [BitServer-4] WARN o.a.d.exec.rpc.ProtobufLengthDecoder - Failure allocating buffer on incoming stream due to memory limits. Current Allocation: 92340224.
> 2018-03-12 11:52:35,929 [BitServer-3] WARN o.a.d.exec.rpc.ProtobufLengthDecoder - Failure allocating buffer on incoming stream due to memory limits. Current Allocation: 92340224.
> 2018-03-12 11:52:35,930 [BitServer-3] ERROR o.a.drill.exec.rpc.data.DataServer - Out of memory in RPC layer.
> 2018-03-12 11:52:35,930 [BitServer-4] ERROR o.a.drill.exec.rpc.data.DataServer - Out of memory in RPC layer.
> 2018-03-12 11:52:35,930 [BitServer-4] WARN o.a.d.exec.rpc.ProtobufLengthDecoder - Failure allocating buffer on incoming stream due to memory limits. Current Allocation: 83886080.
> 2018-03-12 11:52:35,930 [BitServer-3] WARN o.a.d.exec.rpc.ProtobufLengthDecoder - Failure allocating buffer on incoming stream due to memory limits. Current Allocation: 83886080.
> 2018-03-12 11:52:35,930 [BitServer-4] ERROR o.a.drill.exec.rpc.data.DataServer - Out of memory in RPC layer.
> 2018-03-12 11:52:35,930 [BitServer-3] ERROR o.a.drill.exec.rpc.data.DataServer - Out of memory in RPC layer.
> 2018-03-12 11:52:35,931 [BitServer-3] WARN o.a.d.exec.rpc.ProtobufLengthDecoder - Failure allocating buffer on incoming stream due to memory limits. Current Allocation: 83886080.
> 2018-03-12 11:52:35,931 [BitServer-4] WARN o.a.d.exec.rpc.ProtobufLengthDecoder - Failure allocating buffer on incoming stream due to memory limits. Current Allocation: 83886080.
> 2018-03-12 11:52:35,931 [BitServer-3] ERROR o.a.drill.exec.rpc.data.DataServer - Out of memory in RPC layer.
> 2018-03-12 11:52:35,931 [BitServer-4] ERROR o.a.drill.exec.rpc.data.DataServer - Out of memory in RPC layer.
> ...
> ...
> 2018-03-12 11:52:35,939 [BitServer-4] WARN o.a.d.exec.rpc.ProtobufLengthDecoder - Failure allocating buffer on incoming stream due to memory limits. Current Allocation: 67174400.
> 2018-03-12 11:52:35,939 [BitServer-4] ERROR o.a.drill.exec.rpc.data.DataServer - Out of memory in RPC layer.
> 2018-03-12 11:52:35,939 [BitServer-2] WARN o.a.d.exec.rpc.ProtobufLengthDecoder - Failure allocating buffer on incoming stream due to memory limits. Current Allocation: 84017152.
> 2018-03-12 11:52:35,939 [BitServer-2] ERROR o.a.drill.exec.rpc.data.DataServer - Out of memory in RPC layer.
> 2018-03-12 11:52:35,940 [25593391-512d-23ab-7c84-3651006931e2:frag:1:38] INFO o.a.d.e.w.fragment.FragmentExecutor - User Error Occurred: One or more nodes ran out of memory while executing the query. (Failure allocating buffer.)
> org.apache.drill.common.exceptions.UserException: RESOURCE ERROR: One or more nodes ran out of memory while executing the query.
> Failure allocating buffer.
> [Error Id: cf4fd273-d8a2-45e8-8d72-15c738e53b0f ]
>  at org.apache.drill.common.exceptions.UserException$Builder.build(UserException.java:633) ~[drill-common-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.work.fragment.FragmentExecutor.run(FragmentExecutor.java:243) [drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.common.SelfCleaningRunnable.run(SelfCleaningRunnable.java:38) [drill-common-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) [na:1.8.0_161]
>  at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) [na:1.8.0_161]
>  at java.lang.Thread.run(Thread.java:748) [na:1.8.0_161]
> Caused by: org.apache.drill.exec.exception.OutOfMemoryException: Failure allocating buffer.
>  at io.netty.buffer.PooledByteBufAllocatorL.allocate(PooledByteBufAllocatorL.java:67) ~[drill-memory-base-1.13.0-SNAPSHOT.jar:4.0.48.Final]
>  at org.apache.drill.exec.memory.AllocationManager.<init>(AllocationManager.java:83) ~[drill-memory-base-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.memory.BaseAllocator.bufferWithoutReservation(BaseAllocator.java:258) ~[drill-memory-base-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.memory.BaseAllocator.buffer(BaseAllocator.java:241) ~[drill-memory-base-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.memory.BaseAllocator.buffer(BaseAllocator.java:211) ~[drill-memory-base-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.vector.BigIntVector.allocateBytes(BigIntVector.java:239) ~[vector-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.vector.BigIntVector.allocateNew(BigIntVector.java:219) ~[vector-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.vector.RepeatedBigIntVector.allocateNew(RepeatedBigIntVector.java:272) ~[vector-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.vector.AllocationHelper.allocatePrecomputedChildCount(AllocationHelper.java:41) ~[vector-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.vector.AllocationHelper.allocate(AllocationHelper.java:66) ~[vector-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.record.RecordBatchSizer$ColumnSize.allocateVector(RecordBatchSizer.java:403) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.physical.impl.flatten.FlattenRecordBatch.doAlloc(FlattenRecordBatch.java:276) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.physical.impl.flatten.FlattenRecordBatch.handleRemainder(FlattenRecordBatch.java:240) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.physical.impl.flatten.FlattenRecordBatch.innerNext(FlattenRecordBatch.java:165) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.record.AbstractRecordBatch.next(AbstractRecordBatch.java:164) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.physical.impl.BaseRootExec.next(BaseRootExec.java:105) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.physical.impl.SingleSenderCreator$SingleSenderRootExec.innerNext(SingleSenderCreator.java:93) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.physical.impl.BaseRootExec.next(BaseRootExec.java:95) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.work.fragment.FragmentExecutor$1.run(FragmentExecutor.java:233) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at org.apache.drill.exec.work.fragment.FragmentExecutor$1.run(FragmentExecutor.java:226) ~[drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  at java.security.AccessController.doPrivileged(Native Method) ~[na:1.8.0_161]
>  at javax.security.auth.Subject.doAs(Subject.java:422) ~[na:1.8.0_161]
>  at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1595) ~[hadoop-common-2.7.0-mapr-1707.jar:na]
>  at org.apache.drill.exec.work.fragment.FragmentExecutor.run(FragmentExecutor.java:226) [drill-java-exec-1.13.0-SNAPSHOT.jar:1.13.0-SNAPSHOT]
>  ... 4 common frames omitted
> Caused by: io.netty.util.internal.OutOfDirectMemoryError: failed to allocate 67117056 byte(s) of direct memory (used: 8548425728, max: 8589934592)
>  at io.netty.util.internal.PlatformDependent.incrementMemoryCounter(PlatformDependent.java:510) ~[netty-common-4.0.48.Final.jar:4.0.48.Final]
>  at io.netty.util.internal.PlatformDependent.allocateDirectNoCleaner(PlatformDependent.java:464) ~[netty-common-4.0.48.Final.jar:4.0.48.Final]
>  at io.netty.buffer.UnpooledUnsafeNoCleanerDirectByteBuf.allocateDirect(UnpooledUnsafeNoCleanerDirectByteBuf.java:30) ~[netty-buffer-4.0.48.Final.jar:4.0.48.Final]
>  at io.netty.buffer.UnpooledByteBufAllocator$InstrumentedUnpooledUnsafeNoCleanerDirectByteBuf.allocateDirect(UnpooledByteBufAllocator.java:169) ~[netty-buffer-4.0.48.Final.jar:4.0.48.Final]
>  at io.netty.buffer.UnpooledUnsafeDirectByteBuf.<init>(UnpooledUnsafeDirectByteBuf.java:67) ~[netty-buffer-4.0.48.Final.jar:4.0.48.Final]
>  at io.netty.buffer.UnpooledUnsafeNoCleanerDirectByteBuf.<init>(UnpooledUnsafeNoCleanerDirectByteBuf.java:25) ~[netty-buffer-4.0.48.Final.jar:4.0.48.Final]
>  at io.netty.buffer.UnpooledByteBufAllocator$InstrumentedUnpooledUnsafeNoCleanerDirectByteBuf.<init>(UnpooledByteBufAllocator.java:164) ~[netty-buffer-4.0.48.Final.jar:4.0.48.Final]
>  at io.netty.buffer.UnpooledByteBufAllocator.newDirectBuffer(UnpooledByteBufAllocator.java:73) ~[netty-buffer-4.0.48.Final.jar:4.0.48.Final]
>  at io.netty.buffer.AbstractByteBufAllocator.directBuffer(AbstractByteBufAllocator.java:179) ~[netty-buffer-4.0.48.Final.jar:4.0.48.Final]
>  at io.netty.buffer.PooledByteBufAllocatorL$InnerAllocator.newDirectBufferL(PooledByteBufAllocatorL.java:157) ~[drill-memory-base-1.13.0-SNAPSHOT.jar:4.0.48.Final]
>  at io.netty.buffer.PooledByteBufAllocatorL$InnerAllocator.directBuffer(PooledByteBufAllocatorL.java:201) ~[drill-memory-base-1.13.0-SNAPSHOT.jar:4.0.48.Final]
>  at io.netty.buffer.PooledByteBufAllocatorL.allocate(PooledByteBufAllocatorL.java:65) ~[drill-memory-base-1.13.0-SNAPSHOT.jar:4.0.48.Final]
>  ... 27 common frames omitted
> {noformat}



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)