DB2 FOR LUW / DB2FORLUW / 950 - DB2

In certain DPF configurations where a node failure error report 
is sent back to the coord agent, and 
the client application becomes unresponsive, the coord agent may 
get into a state where it can 
not be forced. The node failure report would mark the agent with 
an interrupt of higher priority 
than a force interrupt and hence the force will not move the 
coord agent. In normal sitituations the 
client application should issue a subsequent request, and this 
should drive node failure completion 
and reset the node failure state of the coord agent, after which 
a force should interrupt the coord 
agent. 
 
A coordinator agent in this situation may hang with the 
following stack: 
 
recv 
tcprecv 
sqlcctcprecv 
sqlccrecv 
sqljsSqlam 
sqljsDriveRequests 
sqljsDrdaAsInnerDriver 
sqljsDrdaAsDriver 
RunEDU 
 
And if its interrupt type is: 
 
intrptType: 67108869    <<=== 0x4000005 == 
SQLE_PDB_INTRPT_NODEFAILREP

COORDINATOR AGENT MAY BLOCK ON TCPRECV WAITING FOR REMOTE 
CLIENT 
RESPONSE AND CAN NOT BE FORCED.

DB2 Version 9.5 Fix Pack 9 for Linux, UNIX, and Windows
 DB2 Version 9.5 Fix Pack 10 for Linux, UNIX, and Windows
 

Fix included in v9.5 fixpack 9

keiner bekannt / siehe Local-Fix

Datum - Problem gemeldet    :
Datum - Problem geschlossen :
Datum - der letzten Änderung:

14.06.2011
09.04.2012
09.04.2012

9.5.FP9