Tuning Profile Delivery Mechanism for Linux
From the performance team feedback, the C1 should be used by default and the C0 let to be optional. Signed-off-by: Jaroslav Škarvada <jskarvad@redhat.com> |
||
|---|---|---|
| contrib | ||
| doc | ||
| ktune | ||
| libexec | ||
| monitorplugins | ||
| tests | ||
| tune-profiles | ||
| tuningplugins | ||
| udev | ||
| .gitignore | ||
| AUTHORS | ||
| ChangeLog | ||
| COPYING | ||
| git-daemon-export-ok | ||
| INSTALL | ||
| Makefile | ||
| NEWS | ||
| README | ||
| README.command | ||
| tuned | ||
| tuned-adm | ||
| tuned-adm.consolehelper | ||
| tuned-adm.pam | ||
| tuned.bash | ||
| tuned.conf | ||
| tuned.initscript | ||
| tuned.py | ||
| tuned.service | ||
| tuned.spec | ||
| tuned.tmpfiles | ||
| tuned_adm.py | ||
| tuned_logging.py | ||
| tuned_nettool.py | ||
tuned: A simple daemon that performs monitoring and adaptive configuration
of devices in the system
Copyright (C) 2008, 2009 Red Hat, Inc.
Authors: Phil Knirsch
This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
How to use
----------
After installing the package you simply need to start the tuned
service either with system-config-services or from the
commandline in a rootshell via:
service tuned start
You can also automatically start it during system startup by
switching it on via chkconfig:
chkconfig tuned on
After that the daemon will run automatically and monitor the
usage of various system components. Based on that information
components will then be put into lower or higher power saving
modes to adapt to the current usage. Currently plugins for CPU,
ethernet network and and ATA harddisk devices are implemented.
The config file /etc/tuned.conf contains several sections in
INI style. There is always a "main" section which controls
overall options of the daemon. Currently only 1 option
is available:
interval: Interval for monitoring and tuning in seconds.
Default is 10.
Each monitoring and tuning plugin has an own section then.
Currently only 1 option for plugins is available:
enabled: Enabled or disable the plugin. Default is True.
Any other value disables it.
Monitoring/tuning plugins in detail
-----------------------------------
- CPU
This plugin pair monitors the CPU usage of the system. If the current
workload of it is very low (uptime smaller than 0.2) it will switch
the system wide PM-QOS setting to 999 usec. If the load should get
higher again the PM-QOS latency will be lowered to 100 again. More details
about PM-QOS can be found here http://www.lesswatts.org/projects/power-qos/
- Disk
This plugin pair monitors usage of every disk in the system. For each
harddisk we check how long they have been in idle mode. The longer they are
the more we raise the power saving features for that drive and allow it to
actually spin down. As soon as it is being used we then reduce the power
savings for it and deactivate the automatic spin down again at higher
performance levels again.
- Net
This plugin pair monitors usage of network devices. At startup it checks
what speed the interfaces have. If at any time the utilization of any
interface falls below 5% of the possible maximum speed we then renegotiate
the speed of the interface at 100mbit. In case utilization goes up again
we renegotiate to the full speed once more.