Bug 1020743

Summary:	Update tuned to v2.3
Product:	Red Hat Enterprise Linux 7	Reporter:	Jaroslav Škarvada <jskarvad>
Component:	tuned	Assignee:	Jaroslav Škarvada <jskarvad>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Branislav Blaškovič <bblaskov>
Severity:	unspecified	Docs Contact:
Priority:	unspecified
Version:	7.0	CC:	anton, bblaskov, jeder, jscotka, jskarvad, perfbz
Target Milestone:	beta	Keywords:	Rebase
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:	tuned-2.3.0-1.el7	Doc Type:	Rebase: Bug Fixes and Enhancements
Doc Text:	Rebase package(s) to version: 2.3.0 Highlights, important fixes, or notable enhancements: - various plugins fixes and enhancements, especially regarding performance - documentation updated to be up-to-date with the code - switched to lightweighted pygobject3-base - lowered usage of system resources - fixed several bugs that could lead to the crash	Story Points:	---
Clone Of:		Environment:
Last Closed:	2014-06-13 11:17:27 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:

Description Jaroslav Škarvada 2013-10-18 08:43:12 UTC

Description of problem:
Tuned v2.3 includes various fixes and improvements which are worth to have in RHEL-7 (e.g. fixes for crashes, additional tunings for performance improvements, etc.). Some of the patches are already back-ported and we can back-port the rest, but having the upstream code there would simplify the future maintenance.

The risk of the update is very low, because I am also tuned upstream and I have full control on the content of the update.

Version-Release number of selected component (if applicable):
tuned-2.2.2-4

How reproducible:
Always

Steps to Reproduce:
1. rpm -qv

Actual results:
tuned-2.2.2-4


Expected results:
tuned-2.3.0-1


Additional info:

Comment 10 Branislav Blaškovič 2013-11-07 13:04:35 UTC

I've modified the script to track the times..

#!/bin/bash

PROFILES=(balanced desktop latency-performance powersave sap throughput-performance virtual-guest virtual-host)

for (( i=0; i < 100; i++ ))
do
  p=${PROFILES[$(( RANDOM % ${#PROFILES[@]} ))]}
  echo `date` ": iter: $i, profile: $p"
  tuned-adm profile $p
  echo `date` ": profile switched"
  sleep $(( ( RANDOM % 10 )  + 1 ))
done

# ./tuned-stress 
Thu Nov 7 13:54:28 CET 2013 : iter: 0, profile: virtual-host
Thu Nov 7 13:54:29 CET 2013 : profile switched
Thu Nov 7 13:54:30 CET 2013 : iter: 1, profile: virtual-guest
Thu Nov 7 13:54:39 CET 2013 : profile switched
Thu Nov 7 13:54:49 CET 2013 : iter: 2, profile: sap
Thu Nov 7 13:54:59 CET 2013 : profile switched
Thu Nov 7 13:55:01 CET 2013 : iter: 3, profile: balanced
Thu Nov 7 13:55:09 CET 2013 : profile switched
Thu Nov 7 13:55:18 CET 2013 : iter: 4, profile: virtual-guest
Thu Nov 7 13:55:19 CET 2013 : profile switched
Thu Nov 7 13:55:20 CET 2013 : iter: 5, profile: virtual-host
Thu Nov 7 13:55:29 CET 2013 : profile switched
Thu Nov 7 13:55:34 CET 2013 : iter: 6, profile: balanced
Thu Nov 7 13:55:39 CET 2013 : profile switched
Thu Nov 7 13:55:41 CET 2013 : iter: 7, profile: virtual-host
Thu Nov 7 13:55:49 CET 2013 : profile switched
Thu Nov 7 13:55:50 CET 2013 : iter: 8, profile: sap
Thu Nov 7 13:55:59 CET 2013 : profile switched
Thu Nov 7 13:56:06 CET 2013 : iter: 9, profile: sap
Thu Nov 7 13:56:09 CET 2013 : profile switched
Thu Nov 7 13:56:18 CET 2013 : iter: 10, profile: balanced
Thu Nov 7 13:56:20 CET 2013 : profile switched
Thu Nov 7 13:56:26 CET 2013 : iter: 11, profile: desktop
Thu Nov 7 13:56:30 CET 2013 : profile switched
Thu Nov 7 13:56:34 CET 2013 : iter: 12, profile: virtual-guest
Thu Nov 7 13:56:40 CET 2013 : profile switched
Thu Nov 7 13:56:42 CET 2013 : iter: 13, profile: balanced
Thu Nov 7 13:56:50 CET 2013 : profile switched
Thu Nov 7 13:56:59 CET 2013 : iter: 14, profile: virtual-guest
Thu Nov 7 13:57:00 CET 2013 : profile switched
Thu Nov 7 13:57:10 CET 2013 : iter: 15, profile: desktop
Thu Nov 7 13:57:20 CET 2013 : profile switched
Thu Nov 7 13:57:30 CET 2013 : iter: 16, profile: sap
Thu Nov 7 13:57:40 CET 2013 : profile switched
Thu Nov 7 13:57:49 CET 2013 : iter: 17, profile: powersave
Thu Nov 7 13:57:50 CET 2013 : profile switched
Thu Nov 7 13:57:53 CET 2013 : iter: 18, profile: sap
Thu Nov 7 13:58:00 CET 2013 : profile switched
Thu Nov 7 13:58:03 CET 2013 : iter: 19, profile: virtual-guest
Thu Nov 7 13:58:10 CET 2013 : profile switched
Thu Nov 7 13:58:17 CET 2013 : iter: 20, profile: latency-performance
Thu Nov 7 13:58:21 CET 2013 : profile switched
Thu Nov 7 13:58:23 CET 2013 : iter: 21, profile: powersave
Thu Nov 7 13:58:31 CET 2013 : profile switched
Thu Nov 7 13:58:38 CET 2013 : iter: 22, profile: throughput-performance
Thu Nov 7 13:58:41 CET 2013 : profile switched
Thu Nov 7 13:58:51 CET 2013 : iter: 23, profile: virtual-guest
Thu Nov 7 13:59:01 CET 2013 : profile switched
Thu Nov 7 13:59:05 CET 2013 : iter: 24, profile: sap
Thu Nov 7 13:59:11 CET 2013 : profile switched
Thu Nov 7 13:59:13 CET 2013 : iter: 25, profile: virtual-host
Thu Nov 7 13:59:21 CET 2013 : profile switched
Thu Nov 7 13:59:23 CET 2013 : iter: 26, profile: virtual-host
Thu Nov 7 13:59:31 CET 2013 : profile switched
Thu Nov 7 13:59:32 CET 2013 : iter: 27, profile: sap
Thu Nov 7 13:59:41 CET 2013 : profile switched
Thu Nov 7 13:59:50 CET 2013 : iter: 28, profile: virtual-host
Thu Nov 7 13:59:51 CET 2013 : profile switched
Thu Nov 7 13:59:54 CET 2013 : iter: 29, profile: latency-performance
Thu Nov 7 14:00:02 CET 2013 : profile switched
Thu Nov 7 14:00:07 CET 2013 : iter: 30, profile: sap
Thu Nov 7 14:00:12 CET 2013 : profile switched

It looks ok on my machine. 10 sec time hang was described in comment #7.

Comment 12 Jaroslav Škarvada 2013-11-07 16:31:34 UTC

I am able to reproduce it now, this is the reproducer:

PROFILES=(balanced desktop latency-performance powersave sap throughput-performance virtual-guest virtual-host)

for (( i=0; i < 100; i++ ))
do
  p=${PROFILES[$(( RANDOM % ${#PROFILES[@]} ))]}
  echo "iter: $i, profile: $p"
  tuned-adm profile $p
done

Comment 13 Jaroslav Škarvada 2013-11-07 17:19:32 UTC

This doesn't seem to be regression. I am able to reproduce this with tuned-2.2.2-3 as well.

Comment 14 Jaroslav Škarvada 2013-11-07 17:33:17 UTC

I filled the following bugs:

1) bug 1028122
This is probably not blocker

2) bug 1028119
The source of this problem is currently unclear to me. I am working on it

Comment 15 Jaroslav Škarvada 2013-11-07 22:53:37 UTC

(In reply to Jaroslav Škarvada from comment #14)
> 2) bug 1028119
> The source of this problem is currently unclear to me. I am working on it
Patch with the proposed fix attached.

Comment 16 Branislav Blaškovič 2013-11-19 10:29:43 UTC

Bug 1028122 and bug 1028119 are verified. No other issues found. 
Marking this update bug as verified.

Comment 17 Ludek Smid 2014-06-13 11:17:27 UTC

This request was resolved in Red Hat Enterprise Linux 7.0.

Contact your manager or support representative in case you have further questions about the request.