|
Problem IT16615
|
Status: Closed |
WHEN INTRA_PARALLEL IS ON QUERIES USING HASH JOINS MAY HANG AND NOT BE ABLE TO BE FORCED OFF |
| product: |
DB2 FOR LUW / DB2FORLUW / A50 - DB2 |
| Problem description: |
Systems using INTRA_PARALLEL=YES may see an agent stuck in
'Application status' = UOW Executing.
The stack of this coordinator agent may show:
sqlrr_fetch_error
sqlrr_open
sqljsParseRdbAccessed
All its subagents are in sqloXlatchConflict and at least one
will show a nested sqlri_hsjnFlushBlocks call meaning it dead
locks with itself:
sqloXlatchConflict
sqlri_hsjnFlushBlocks...
sqlri_hsjnHeapReduction...
sqlri_hsjnGetTupleBlock...
sqlri_hsjnFlushBlocks...
...
db2pd -latches shows all subagents waiting for the same latch
SQLO_LT_sqlriSectInit_C__dynamic_xlatch, but there is no holder
of this latch:
Latch Waiters With No Holders:
Address Holder Waiter Filename
LOC LatchType
0x0A000100117FAC00 0 97631 sqlktins.C 50
SQLO_LT_sqlriSectInit_C__dynamic_xlatch
0x0A000100117FAB80 0 97888 sqlktins.C 50
SQLO_LT_sqlriSectInit_C__dynamic_xlatch
0x0A000100117FAC00 0 98145 sqlktins.C 50
SQLO_LT_sqlriSectInit_C__dynamic_xlatch
0x0A000100117FAC00 0 98402 sqlktins.C 50
SQLO_LT_sqlriSectInit_C__dynamic_xlatch
Trying to force this application does not work. It keeps
running.
If this application was doing an insert/update/delete operation,
you may run into of log_full situation.
The only way out of this situation is to kill the db2 instance. |
| Problem Summary: |
****************************************************************
* USERS AFFECTED: *
* ALL *
****************************************************************
* PROBLEM DESCRIPTION: *
* See Error Description *
****************************************************************
* RECOMMENDATION: *
* Upgrade to Db2 10.5 Fix Pack 9 or higher *
**************************************************************** |
| Local Fix: |
Use a fixed value for sortheap, instead of letting STMM tuning
it.
Check db2diag.log for sortheap tuning entries and chose a value
at the upper end in order to avoid having too low sortheap
memory. |
| Solution |
First fixed in Db2 10.5 Fix Pack 9 |
| Workaround |
not known / see Local fix |
| Timestamps |
Date - problem reported :
Date - problem closed :
Date - last modified :
| 16.08.2016
29.09.2017
29.09.2017 |
| Problem solved at the following versions (IBM BugInfos) |
9.0.
|
| Problem solved according to the fixlist(s) of the following version(s) |
