|
Problem IC81329
|
Status: Closed |
DB2 LOAD MAY HANG WHEN MESSAGE QUEUE SPACE HAS BEEN EXHAUSTED |
| product: |
DB2 FOR LUW / DB2FORLUW / 970 - DB2 |
| Problem description: |
On Linux, Solaris, or HP-UX a DB2 LOAD may hang when the
kernel's message queue space has been exhausted. You may
observe the LIST UTILITIES shows the completed work is equal to
total work but the LOAD has not completed. For example:
$ db2 list utilities show detail
ID = 1179
Type = LOAD
Database Name = S232
Partition Number = 0
Description = [LOAD COMMAND]
Start Time = 02/02/2012 02:02:02.000000
State = Executing
Invocation Type = User
Progress Monitoring:
Phase Number = 1
Description = SETUP
Total Work = 0 bytes
Completed Work = 0 bytes
Start Time = 02/02/2012 02:02:02.000000
Phase Number [Current] = 2
Description = LOAD
Total Work = 2000000000 rows
Completed Work = 2000000000 rows
Start Time = 02/02/2012 02:02:03.000000
You may also observe a stack for the db2agent executing the LOAD
that looks like:
msgrcv
sqlorqueInternal
sqlorque2
sqlulRead
...
sqluvload
The stack for the db2lbm (load buffer manipulator) may show:
msgsnd
sqlowqueInternal
sqlowqueWrapper
sqlowqueNoBlock
sqluWriteQueue
Inspecting "ipcs -q" output can also reveal symptoms of this
hang. If the used-bytes field of any message queue is near or
at the limit of the system configured queue size (64K in this
example), then this is an indication that the message queue
space has been exhausted.
------ Message Queues --------
key msqid owner perms used-bytes
messages
0x00000000 2047508665 leechu 601 5248 16
0x00000000 2047410358 leechu 601 7744 138
0x00000000 2047344819 leechu 601 25088 511
0x00000000 2047377589 leechu 601 65496 1248 |
| Problem Summary: |
****************************************************************
* USERS AFFECTED: *
* ALL *
****************************************************************
* PROBLEM DESCRIPTION: *
* See Error Description *
****************************************************************
* RECOMMENDATION: *
* Upgrade to DB2 Version 9.7 Fix Pack 6 *
**************************************************************** |
| Local Fix: |
Increasing the kernel setting for message queue size (MSGMNB for
Linux) should avoid the problem in most cases. Decreasing
UTIL_HEAP_SZ, CPU/DISK PARALLELISM, or specifying a smaller DATA
BUFFER SIZE may also help. |
| available fix packs: |
DB2 Version 9.7 Fix Pack 6 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 7 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 8 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 9 for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 9a for Linux, UNIX, and Windows
DB2 Version 9.7 Fix Pack 10 for Linux, UNIX, and Windows
|
| Solution |
First fixed in Version 9.7 Fix Pack 6 |
| Workaround |
not known / see Local fix |
| Timestamps |
Date - problem reported :
Date - problem closed :
Date - last modified :
| 09.02.2012
10.12.2012
10.12.2012 |
| Problem solved at the following versions (IBM BugInfos) |
9.7.FP6
|
| Problem solved according to the fixlist(s) of the following version(s) |
| 9.7.0.6
|
 |
