DB2 FOR LUW / DB2FORLUW / A50 - DB2

In the case where a GPFS mount path is a substring of another 
mount path on the system, the mountV105_monitor.ksh script may 
erroneously report the GPFS mount resource as "Stuck Online" 
instead of "Online". This issue only exists on AIX. 
 
For example, take the following two filesystem entries from the 
/etc/filesystems file: 
------------------------- 
/opt/IBM/db2: 
        dev             = /dev/db2lv 
        vfs             = jfs2 
        log             = INLINE 
        mount           = true 
        account         = false 
 
/db2: 
        dev             = /dev/db2 
        vfs             = mmfs 
        nodename        = - 
        mount           = mmfs 
        type            = mmfs 
        account         = false 
        options         = rw,mtime,atime,dev=db2 
------------------------- 
 
When adding the /db2 GPFS mount path to the TSA resource model, 
the following error is seen in the mmfs.log file: 
Thu Aug 20 14:40:46 HKT 2015: mountV105_monitor.ksh[17694736]: 
Automounting is enable for /db2. Please change this. 
 
The following error message is also seen in the db2diag.log: 
-------------------------------- 
2015-08-20-15.48.07.993625+480 I1635499A382         LEVEL: 
Warning 
PID     : 10747950             TID : 1              PROC : 
db2instance 
INSTANCE: db2inst1              NODE : 000 
HOSTNAME: DB2HOST01 
EDUID   : 1 
FUNCTION: DB2 UDB, high avail services, 
sqlhaCheckMountResourcesOnHost, probe:4113 
DATA #1 : <preformatted> 
The mount db2mnt-db2-rs is not online on host DB2HOST02 
 
Furthermore, the lssam command output will show the resource 
group in the following state: 
Stuck online IBM.ResourceGroup:db2mnt-db2-rg Nominal=Online 
        '- Stuck online IBM.Application:db2mnt-db2-rs 
                |- Stuck online 
IBM.Application:db2mnt-db2-rs:DB2HOST01 
                '- Stuck online 
IBM.Application:db2mnt-db2-rs:DB2HOST02 
 
This is all due to a logic error with the following line taken 
from the mountV105_monitor.ksh script: 
fstype=$(cat /etc/filesystems | grep -p ${MP?}: | awk '{print $1 
" " $2 " " $3}' | egrep "^vfs = " | awk '{print $3}') 
 
The '^' character should be added to the above line as shown 
below in order to correct this issue: 
fstype=$(cat /etc/filesystems | grep -p ^${MP?}: | awk '{print 
$1 " " $2 " " $3}' | egrep "^vfs = " | awk '{print $3}')

**************************************************************** 
* USERS AFFECTED:                                              * 
* ALL                                                          * 
**************************************************************** 
* PROBLEM DESCRIPTION:                                         * 
* See Error Description                                        * 
**************************************************************** 
* RECOMMENDATION:                                              * 
* Upgrade to V10.5 fp7                                         * 
****************************************************************

Make the following change to the 
/usr/sbin/rsct/sapolicies/db2/mountV105_monitor.ksh file on all 
cluster nodes: 
Change: 
fstype=$(cat /etc/filesystems | grep -p ${MP?}: | awk '{print $1 
" " $2 " " $3}' | egrep "^vfs = " | awk '{print $3}') 
To: 
fstype=$(cat /etc/filesystems | grep -p ^${MP?}: | awk '{print 
$1 " " $2 " " $3}' | egrep "^vfs = " | awk '{print $3}')

not known / see Local fix

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

09.09.2015
20.01.2016
20.01.2016