Hi Gerard,
We have been using these kernel parameters for internal accounts. It is
based on limit of 90% for DB2:
#!/bin/bash
# Configuring kernel parameters on /etc/sysctl.conf
shmmax=$(echo (($(cat /proc/meminfo | grep ^MemTotal | awk '{print
$2}')*90)/100)*1024 | bc)
shmall=$(echo $shmmax/4096 | bc)
sem=250 256000 32 1024
msgmni=1024
msgmax=65536
msgmnb=65536
rmem_default=16777216
wmem_default=16777216
rmem_max=16777216
wmem_max=16777216
tcp_rmem=4096 87380 16777216
tcp_wmem=4096 65536 16777216
swappiness=0
pagecluster=1
overcommit_ratio=75
min_free_kbytes=$(echo (($(cat /proc/meminfo | grep ^MemTotal | awk
'{print $2}')*5)/100) | bc)
dirty_ratio=20
dirty_background_ratio=5
#-
#| Memory
Settings
|
#-
kernel.shmmax=${shmmax}
kernel.shmall=${shmall}
kernel.sem=${sem}
kernel.msgmni=${msgmni}
kernel.msgmax=${msgmax}
kernel.msgmnb=${msgmnb}
vm.swappiness=${swappiness}
vm.page-cluster=${pagecluster}
vm.overcommit_ratio=${overcommit_ratio}
vm.min_free_kbytes=${min_free_kbytes}
vm.dirty_ratio=${dirty_ratio}
vm.dirty_background_ratio=${dirty_background_ratio}
#-
#| Network
Settings
|
#-
net.core.rmem_default=${rmem_default}
net.core.wmem_default=${wmem_default}
net.core.rmem_max=${rmem_max}
net.core.wmem_max=${wmem_max}
net.ipv4.tcp_rmem=${tcp_rmem}
net.ipv4.tcp_wmem=${tcp_wmem}
Good luck!
Tito Garrido Ogando
On Thu, May 7, 2015 at 5:10 PM, Gerard Howells ghowe...@americafirst.com
wrote:
Greetings VM List!
We recently moved a database from DB2 9.5 running on SLES10 SP4 to DB2
10.5 (fix pack 5) running on SLES11 SP3, both under z/VM 6.2. We're having
some issues with increased memory consumption and wondered if anybody else
had noticed a similar trend. To give a little background the old guest ran
wonderfully efficiently for a number of years with 2G of memory allocated
to the guest. The new guest has been sized up through the testing period to
8G and we're still having problems. The recommended mode for DB2's memory
management is to run with its automatic memory management feature enabled.
This seems, however, to just gobble up all the memory we can give it and
has led to the guest freezing until we remove the load from the application
running queries, at which time it becomes responsive again. Any advice on
either OS configuration parameters (beyond what IBM recommends here:
http://www-01.ibm.com/support/knowledgecenter/SSEPGG_10.5.0/com.ibm.db2.luw.qb.server.doc/doc/c0057140.html?cp=SSEPGG_10.5.0%2F2-0-1-2-2-0-10-1)
to be aware of or DB2 configuration that might help us with this would be
enormously appreciated.
Thanks all,
Gerard Howells
Mainframe Systems Programmer
Enterprise Systems
America First Credit Union
TEL: 801-827-8353tel:801-827-8353
ghowe...@americafirst.commailto:ghowe...@americafirst.com
[America First Credit Union]
https://www.americafirst.com/redirect/clickthrough.cfm?dest=exEmail
Notice: This communication and any attachments are intended for a specific
individual and purpose. The information contained in this communication is
confidential and protected by law. If you are not the intended recipient
and received this communication in error, please notify the sender
immediately by reply e-mail, or call 801-627-0900, and delete all copies.
Any review, retention, or disclosure to others is strictly prohibited.
Thank you.
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or
visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/
--
Linux User #387870
.
_/_õ|__|
..º[ .-.___.-._| . . . .
.__( o)__( o).:___
--
For LINUX-390 subscribe / signoff / archive access instructions,
send email to lists...@vm.marist.edu with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390
--
For more information on Linux on System z, visit
http://wiki.linuxvm.org/