Uploaded image for project: 'IMPALA'
  1. IMPALA
  2. IMPALA-8452

Avro scanner seems broken

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Resolved
    • Major
    • Resolution: Cannot Reproduce
    • Impala 3.3.0
    • None
    • Backend
    • ghx-label-8

    Description

      A few scanner tests started failing recently on Centos6. Coincidentally, both of them only started happening after this commit. attilaj, can you please take a look and reassign if you think that commit is unrelated ?

      Oddly enough, this has shown up on Centos6. Other exhaustive runs with Centos7 seem to work fine. May be it's related to some platform's library ?

      In the first case, a select count star from an avro table hangs for 2 hours:

      query_test/test_scanners_fuzz.py:83: in test_fuzz_alltypes
          self.run_fuzz_test(vector, src_db, table_name, unique_database, table_name)
      query_test/test_scanners_fuzz.py:201: in run_fuzz_test
          result = self.execute_query(query, query_options = query_options)
      common/impala_test_suite.py:619: in wrapper
          return function(*args, **kwargs)
      common/impala_test_suite.py:650: in execute_query
          return self.__execute_query(self.client, query, query_options)
      common/impala_test_suite.py:721: in __execute_query
          return impalad_client.execute(query, user=user)
      common/impala_connection.py:180: in execute
          return self.__beeswax_client.execute(sql_stmt, user=user)
      beeswax/impala_beeswax.py:183: in execute
          handle = self.__execute_query(query_string.strip(), user=user)
      beeswax/impala_beeswax.py:360: in __execute_query
          self.wait_for_finished(handle)
      beeswax/impala_beeswax.py:384: in wait_for_finished
          time.sleep(0.05)
      E   Failed: Timeout >7200s
      
      SET client_identifier=query_test/test_scanners_fuzz.py::TestScannersFuzzing::()::test_fuzz_alltypes[protocol:beeswax|exec_option:{'debug_action':None;'abort_on_error':False;'mem_limit':'512m';'num_nodes':0}|table_format:avro/none];
      SET batch_size=1;
      SET num_nodes=0;
      SET disable_codegen_rows_threshold=0;
      SET disable_codegen=True;
      SET abort_on_error=False;
      SET mem_limit=512m;
      -- executing against localhost:21000
      select count(*) from test_fuzz_alltypes_2cdcb963.alltypes q;
      
      -- 2019-04-24 04:14:31,857 INFO     MainThread: Started query 2049069f9f5e3aa8:f2fd47ff00000000
      

      The second case has to do with incorrect number of rows in a select count star from tpch_avro.lineitem:

      query_test/test_scanners.py:947: in test_tpch_scan_ranges
          self.run_test_case('tpch-scan-range-lengths', vector)
      common/impala_test_suite.py:517: in run_test_case
          self.__verify_results_and_errors(vector, test_section, result, use_db)
      common/impala_test_suite.py:370: in __verify_results_and_errors
          replace_filenames_with_placeholder)
      common/test_result_verifier.py:449: in verify_raw_results
          VERIFIER_MAP[verifier](expected, actual)
      common/test_result_verifier.py:271: in verify_query_result_is_equal
          assert expected_results == actual_results
      E   assert Comparing QueryTestResults (expected vs actual):
      E     6001215 != 6000679
      
      -- 2019-04-24 03:43:42,805 INFO     MainThread: max_scan_range_length=8412307
      SET client_identifier=query_test/test_scanners.py::TestTpchScanRangeLengths::()::test_tpch_scan_ranges[protocol:beeswax|exec_option:{'batch_size':0;'num_nodes':0;'disable_codegen_rows_threshold':0;'disable_codegen':False;'abort_on_error':1;'exec_single_node_rows_threshold':0}|;
      -- executing against localhost:21000
      use tpch_avro;
      
      -- 2019-04-24 03:43:42,814 INFO     MainThread: Started query c04e1968443b52fc:5b99b1b300000000
      SET client_identifier=query_test/test_scanners.py::TestTpchScanRangeLengths::()::test_tpch_scan_ranges[protocol:beeswax|exec_option:{'batch_size':0;'num_nodes':0;'disable_codegen_rows_threshold':0;'disable_codegen':False;'abort_on_error':1;'exec_single_node_rows_threshold':0}|;
      SET batch_size=0;
      SET num_nodes=0;
      SET disable_codegen_rows_threshold=0;
      SET disable_codegen=False;
      SET abort_on_error=1;
      SET max_scan_range_length=8412307;
      SET exec_single_node_rows_threshold=0;
      -- executing against localhost:21000
      select count(*)
      from lineitem;
      

      Attachments

        Issue Links

          Activity

            People

              Unassigned Unassigned
              kwho Michael Ho
              Votes:
              0 Vote for this issue
              Watchers:
              3 Start watching this issue

              Dates

                Created:
                Updated:
                Resolved: