DB2 FOR LUW / DB2FORLUW / B10 - DB2

db2sysc sends SIGKILL to itself which generates FODC_Panic
diagnostics when stopping nodes concurrently. This affects
multiple partition systems.

Reproduce by running the script below. This example uses 4
partition system.

#!/bin/sh
db2 "activate db t"
sleep 1
db2stop force nodenum 1 &
db2stop force nodenum 3 &
db2stop force nodenum 2 &
db2stop force nodenum 0 &
sleep 1
db2stop force nodenum 0 &
db2stop force nodenum 3 &
db2stop force nodenum 1 &
db2stop force nodenum 2 &
sleep 1
db2stop force nodenum 3 &
db2stop force nodenum 0 &
db2stop force nodenum 3 &
db2stop force nodenum 1 &
sleep 1
db2stop force nodenum 1 &
db2stop force nodenum 3 &
db2stop force nodenum 0 &
db2stop force nodenum 2 &


NOTES:
a) By default this fix is disabled. To enable this fix, set DB2
registry variable DB2_PMODEL_SETTINGS=PARALLEL_SINGLE_STOP:TRUE
then restart DB2 instance for new change to take effect
b) Can only be enabled on DPF instance which has more than one
member

****************************************************************
* USERS AFFECTED:                                              *
* ALL                                                          *
****************************************************************
* PROBLEM DESCRIPTION:                                         *
* See Error Description                                        *
****************************************************************
* RECOMMENDATION:                                              *
* Upgrade to Db2 11.1 Mod 4 Fixpack 4 or higher                *
****************************************************************

Two alternatives:
a) "db2stop force" parallel db2stop on all nodes, but it lets
DB2 decide order to stop each node
OR
b) Execute "db2stop force on nodenum xx" sequentially instead of
concurrently as per the reproduction script.

See LOCAL FIX.

Date  - problem reported    :
Date  - problem closed      :
Date  - last modified       :

18.07.2018
27.11.2018
03.12.2018