Skip to content
This repository has been archived by the owner on Sep 18, 2023. It is now read-only.

Shuffle split SegmentFault issue on complex type Array, exposed from customer Case #602

Closed
zhixingheyi-tian opened this issue Dec 5, 2021 · 2 comments
Labels
bug Something isn't working

Comments

@zhixingheyi-tian
Copy link
Collaborator

zhixingheyi-tian commented Dec 5, 2021

From customer query, encountered SegmentFault issue.

Backtrace the core dump file:

#6  <signal handler called>
#7  0x00007f8623a13b00 in arrow::Status sparkcolumnarplugin::shuffle::Splitter::AppendList<arrow::ListType, arrow::StringType, arrow::ListArray, arrow::ListBuilder>(std::shared_ptr<arrow::ListArray> const&, std::vector<std::shared_ptr<arrow::ListBuilder>, std::allocator<std::shared_ptr<arrow::ListBuilder> > > const&, long) ()
   from /tmp/spark_columnar_plugin_5046574679083750592/libspark_columnar_jni.so
#8  0x00007f86239fb863 in sparkcolumnarplugin::shuffle::Splitter::SplitListArray(arrow::RecordBatch const&) ()
   from /tmp/spark_columnar_plugin_5046574679083750592/libspark_columnar_jni.so
#9  0x00007f8623a004be in sparkcolumnarplugin::shuffle::Splitter::DoSplit(arrow::RecordBatch const&) ()
   from /tmp/spark_columnar_plugin_5046574679083750592/libspark_columnar_jni.so
#10 0x00007f8623a005f8 in sparkcolumnarplugin::shuffle::Splitter::Split(arrow::RecordBatch const&) ()
   from /tmp/spark_columnar_plugin_5046574679083750592/libspark_columnar_jni.so
#11 0x00007f86236032cf in Java_com_intel_oap_vectorized_ShuffleSplitterJniWrapper_split ()
   from /tmp/spark_columnar_plugin_5046574679083750592/libspark_columnar_jni.so

The input recordbatch is:

in->ToString():col_76332#224:   [
    9047,
    5935
  ]
room_id#220:   [
    311,
    4369
  ]
col_52318#221:   [
    "drmyeitlkt",
    "fjtgvacknt"
  ]
col_3859#234:   [
    5131,
    1329
  ]
os#235:   [
    "tkwgprkbyg",
    "nirdfoawle"
  ]
col_94460#236:   [
    "pdyxckpsdm",
    "oufaqcojnx"
  ]
col_41078#237:   [
    "dihhuzhyco",
    "texdmhhvwb"
  ]
col_37774#238:   [
    "ljldmzkdup",
    "dljsjthjni"
  ]
col_13272#252:   [
    [
      "quphxbgjyx",
      "josfcmuiir",
      "opcomsjqar",
      "fhfdbchyrv",
      "jbzegbnwfj"
    ],
    [
      "rflhvnmymh",
      "izsranbqoz",
      "spjfaqxelp",
      "dswtpjyalm",
      "kwgjhehiya"
    ]
  ]
col_70363#245:   [
    "kfsbezwevo",
    "baglyleljo"
  ]
col_74054#246:   [
    1598803200,
    1598803200
  ]
col_68298#247:   [
    "nzlykrdytv",
    "ozqhtoeknb"
  ]
col_52825#248:   [
    5405,
    4442
  ]
col_84617#254:   [
    3885,
    4635
  ]
col_18854#257:   [
    693,
    807
  ]
col_7492#331:   [
    9409,
    9519
  ]
col_55231#332:   [
    -1598851795,
    -1598852758
  ]
live_id#267:   [
    10,
    10
  ]
app_id#258:   [
    3541,
    9600
  ]


@zhixingheyi-tian
Copy link
Collaborator Author

cc @zhouyuan @zhztheplayer

@zhixingheyi-tian zhixingheyi-tian changed the title Shuffle split SegmentFault issue on complex type Array, exposed from ByteDance Case Shuffle split SegmentFault issue on complex type Array, exposed from ByteDance Case Dec 6, 2021
@zhixingheyi-tian zhixingheyi-tian changed the title Shuffle split SegmentFault issue on complex type Array, exposed from ByteDance Case Shuffle split SegmentFault issue on complex type Array, exposed from customer Case Dec 6, 2021
@zhouyuan
Copy link
Collaborator

close in #623

@haojinIntel haojinIntel added the bug Something isn't working label Apr 8, 2022
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
bug Something isn't working
Projects
None yet
Development

No branches or pull requests

3 participants