A profile can now be set in two modes: automatic and manual.
When the automatic mode is active, Tuned always activates the
recommended profile upon startup. This mode can be set using
the following command:
tuned-adm auto_profile
When the manual mode is active, the previously applied profile
is activated upon startup. This mode can be set by switching
to the desired profile, e.g.
tuned-adm profile throughput-performance
The mode is saved on the second line of /etc/tuned/active_profile.
When the file /etc/tuned/active_profile does not exist, or is
empty, automatic mode is enabled. When the file has only one line,
i.e. it was generated by a previous Tuned version, manual mode
is enabled so that the upgrade doesn't cause unexpected active profile
changes.
Related: rhbz#1459146
Signed-off-by: Ondřej Lysoněk <olysonek@redhat.com>
Only 'y', 'yes', 't', 'true' and '1' are accepted as true values
for initrd_remove_dir (case insensitive). Other values are
interpreted as false.
Resolves: rhbz#1455161
Signed-off-by: Ondřej Lysoněk <olysonek@redhat.com>
Tuna is already executed in the realtime profile which is included by
the realtime-virtual-host/guest profiles, there is no need to execute
it twice. Also fixed minor whitespace inconsistency.
Related: rhbz#1472840
Signed-off-by: Jaroslav Škarvada <jskarvad@redhat.com>
Also removed the parameter from the realtime profile, because it already
includes the network-latency profile.
Resolves: rhbz#1451073
Signed-off-by: Jaroslav Škarvada <jskarvad@redhat.com>
Steps to reproduce :
============
1. yum install tuned.
2. yum install tuned-profiles-cpu-partitioning.noarch
3. tuned-adm profile cpu-partitioning
4. tuned-adm verify -i
Verification failed, current system settings differ from the preset profile.
You can mostly fix this by Tuned restart, e.g.:
service tuned restart
Sometimes (if some plugins like bootloader are used) also reboot is required.
See tuned log file ('/var/log/tuned/tuned.log') for details.
Python traces are logged in /var/log/tuned/tuned.log file
INFO tuned.plugins.plugin_script: calling script '/usr/lib/tuned/cpu-partitioning/script.sh' with arguments '['verify', 'ignore_missing']'
ERROR tuned.plugins.plugin_script: script '/usr/lib/tuned/cpu-partitioning/script.sh' error: 1, 'Traceback (most recent call last):
File "/usr/libexec/tuned/defirqaffinity.py", line 102, in <module>
verify(mask)
File "/usr/libexec/tuned/defirqaffinity.py", line 64, in verify
interruptdirs.remove("2")
ValueError: list.remove(x): x not in list'
ERROR tuned.plugins.plugin_script: verify: failed: '[u'/usr/lib/tuned/cpu-partitioning/script.sh']'
If the value 2 is not in the list, call interruptdirs.remove("2") will throw ValueError: list.remove(x): x not in list'.
Fix the code to properly handle ValueError.
Signed-off-by: Seeteena Thoufeek <s1seetee@linux.vnet.ibm.com>
E.g.:
[myprofile]
process=myprocess1
process2=myprocess2.*
The myprofile will be recommended if there are 'myprocess1' and
'myprocess2.*' (e.g. myprocess2, myprocess21, myprocess2_whatever)
running.
Resolves: rhbz#1461838
Signed-off-by: Jaroslav Škarvada <jskarvad@redhat.com>
From 63e5ebfe0c71c5e6d0613c4d8327b2a6f3fba7ba Mon Sep 17 00:00:00 2001
From: Luiz Capitulino <lcapitulino@redhat.com>
Date: Fri, 19 May 2017 14:42:06 -0400
Subject:
The realtime guest profile uses the RT kernel default priority
assignment for the ksoftird, rcuc, and ktimersoftd kernel threads,
which is:
17 FF 2 [rcuc/1] *
19 FF 1 [ktimersoftd/1] *
20 TS - [ksoftirqd/1] *
This doesn't allow running a PMD thread with fifo:1 in the
guest, since this will starve ktimersoftd and ksoftirqd if
those threads ever become runnable.
The solution is to use the same priority assigment we use
in the host where we've been running PMD threads with
fifo:1 priorities without problems. This assignment is:
151 FF 4 [rcuc/15] *
153 FF 3 [ktimersoftd/15] *
154 FF 2 [ksoftirqd/15] *
This patch just copies the [scheduler] section from the
host profile.
Resolves: rhbz#1452357
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
By default the timer tick will fire at the same time on all
CPUs. This causes the jiffies_lock to get contended in the
real-time kernel, as it's a raw spinlock. skew_tick=1 causes
the timer tick to fire at different times on all CPUs,
fixing the contention.
Note skew_tick=1 usage is going to be propagated to the
NFV profiles as well.
Resolves: rhbz#1447938
Signed-off-by: Luiz Capitulino <lcapitulino@redhat.com>
