[IMPALA-2531] partitioned-hash-join-node.cc:233] Check failed: null_probe_rows_ != __null

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: Impala 2.3.0
• Fix Version/s: Impala 2.3.0
• Labels: crash stress
• Environment: impalad version 2.3.0-cdh5-INTERNAL DEBUG (build 58309c22c1b7e4c33cf0744829a658c3123da037); Built on Fri, 09 Oct 2015 22:55:42 PST
• Target Version: Impala 2.3.0
• Assignee: Tim Armstrong
• Reporter: casey
• Created: 11/Oct/15 03:40
• Updated: 16/Nov/15 21:42
• Resolved: 12/Oct/15 05:08

Description

A stress test run crashed after ~3 hours and ~6k queries saying

impala-stress-kerberized-3.vpc.cloudera.com crashed:
F1010 20:17:12.567666  1405 partitioned-hash-join-node.cc:233] Check failed: null_probe_rows_ != __null
[...skipped...]
#6  0x000000000205f92d in google::LogMessageFatal::~LogMessageFatal (this=0x7fa08c29cdb0, __in_chrg=<value optimized out>) at src/logging.cc:1836
#7  0x000000000168412c in impala::PartitionedHashJoinNode::ClosePartitions (this=0x2660b7b00) at /usr/src/debug/impala-2.3.0-cdh5.5.0-SNAPSHOT/be/src/exec/partitioned-hash-join-node.cc:233
#8  0x0000000001684210 in impala::PartitionedHashJoinNode::Close (this=0x2660b7b00, state=0x8a525200) at /usr/src/debug/impala-2.3.0-cdh5.5.0-SNAPSHOT/be/src/exec/partitioned-hash-join-node.cc:247
#9  0x0000000001598a39 in impala::ExecNode::Close (this=0x128ff6300, state=0x8a525200) at /usr/src/debug/impala-2.3.0-cdh5.5.0-SNAPSHOT/be/src/exec/exec-node.cc:179
#10 0x00000000016700e3 in impala::PartitionedAggregationNode::Close (this=0x128ff6300, state=0x8a525200) at /usr/src/debug/impala-2.3.0-cdh5.5.0-SNAPSHOT/be/src/exec/partitioned-aggregation-node.cc:433
#11 0x0000000001562005 in impala::PlanFragmentExecutor::Close (this=0x9b18d428) at /usr/src/debug/impala-2.3.0-cdh5.5.0-SNAPSHOT/be/src/runtime/plan-fragment-executor.cc:573
#12 0x00000000013330d6 in impala::FragmentMgr::FragmentExecState::Exec (this=0x9b18d200) at /usr/src/debug/impala-2.3.0-cdh5.5.0-SNAPSHOT/be/src/service/fragment-exec-state.cc:51
#13 0x000000000132b6d8 in impala::FragmentMgr::FragmentExecThread (this=0x4fc4900, exec_state=0x9b18d200) at /usr/src/debug/impala-2.3.0-cdh5.5.0-SNAPSHOT/be/src/service/fragment-mgr.cc:83
#14 0x000000000132f0c8 in boost::_mfi::mf1<void, impala::FragmentMgr, impala::FragmentMgr::FragmentExecState*>::operator() (this=0x2c50750a0, p=0x4fc4900, a1=0x9b18d200) at /opt/toolchain/boost-pic-1.55.0/include/boost/bind/mem_fn_template.hpp:165
#15 0x000000000132ee81 in boost::_bi::list2<boost::_bi::value<impala::FragmentMgr*>, boost::_bi::value<impala::FragmentMgr::FragmentExecState*> >::operator()<boost::_mfi::mf1<void, impala::FragmentMgr, impala::FragmentMgr::FragmentExecState*>, boost::_bi::list0> (this=0x2c50750b0, f=..., a=...) at /opt/toolchain/boost-pic-1.55.0/include/boost/bind/bind.hpp:313
#16 0x000000000132e769 in boost::_bi::bind_t<void, boost::_mfi::mf1<void, impala::FragmentMgr, impala::FragmentMgr::FragmentExecState*>, boost::_bi::list2<boost::_bi::value<impala::FragmentMgr*>, boost::_bi::value<impala::FragmentMgr::FragmentExecState*> > >::operator() (this=0x2c50750a0) at /opt/toolchain/boost-pic-1.55.0/include/boost/bind/bind_template.hpp:20

The cluster was using kerberos, not sure if that is related.