2839982 - Server may become unresponsive when running a large number of rows INSERT - SAP IQ

• SQL statements running an INSERT consisting of  a large number of rows can cause the SAP IQ server to become unresponsive to other requests. 
  
  
• The SAP IQ server process may end up completely hung and its CPU usage becomes zero.  At this point the process needs to be killed and restarted.
  
  
• This issue had been reported when using client tools and when the number of rows being inserted exceeds 50 000 rows.
  
  
• The following types of message may also get reported in the std error a .srv server log files...

  W. 08/01 07:59:10.200156 Task 0x68d720(Request task 144) is trying get forbid mutex held by task 0x68df00(Request task 146) for more than 60000 ms
  W. 08/01 07:59:10.200743 Task 0x68df00(Request task 146) state: OS thread id=0x7ffdb9749700
  W. 08/01 07:59:10.200967 Task 0x68df00(Request task 146) state: worker=0x7ffdf4000f20 wakend=0 waiting=0
  W. 08/01 07:59:10.201333 Worker 0x7ffdf4000f20 state: wakend=0 waiting=0 in_atomic=0 is_dedicated=0 in_critical_section=0 allow_forbid=0

• Stack traces taken of the hung SAP IQ server process may (but not necessarily always)  show a thread hung on a mutex call during a checkpoint...

  UnixTask::wait
  CondVar::wait(Mutex&)
  TimedMutex::get()
  Worker::get_all_forbid_mutexes()
  Worker::forbid
  DB_Checkpoint
  DbAutoCkpt::force_checkpoint()

  
  
  
• The stack traces taken of the hung process may (but not necessarily always) show one or more threads for new queries stuck during initial query parsing...

  SQLParser::pa_join_plist
  SQLParser::actions
  SQLParser::yyparse()
  PSpawns
  SQLParser::Parse
  ParseSupport::StringToCommand
  
  

• SAP IQ 16.0
• SAP IQ 16.1

hang, sybase, stuck, deadlock, pstack, slow, performance , degrade, java, dbisql, isql, CR819842, CR#1819842, 819842, crash, dbping, freezes, pingable, blocked, records, list, long, bug, values, SQLParser::pa_join_plist , KBA , BC-SYB-IQ , Sybase IQ , Problem