LinuxQuestions.org
Welcome to the most active Linux Forum on the web.
Go Back   LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Debian
User Name
Password
Debian This forum is for the discussion of Debian Linux.

Notices


View Poll Results: systemd vs. upstart, or else?
sysVinit: if they can't decide just keep the status quo, it has worked for 40+ years. 43 58.90%
openrc: a more traditional init system 18 24.66%
upstart: the Canonical non GNU way 3 4.11%
systemd: the RHEL/SuSE/Pottering way. 12 16.44%
Multiple init systems, just let the user decide and leave the nightmare for the maintainers. 10 13.70%
Don't know/don't really care. 6 8.22%
Multiple Choice Poll. Voters: 73. You may not vote on this poll

Reply
  Search this Thread
Old 02-06-2014, 06:31 AM   #16
vl23
Member
 
Registered: Mar 2009
Posts: 125

Original Poster
Rep: Reputation: 8

Quote:
Originally Posted by TobiSGD View Post
I don't know where you got this from, but jails are a completely different thing. You can use cgroups and containers to achieve something similar to jails, but you don't have to use it that way, mostly it is used it to limit resources for specific processes and make it easier to reliably shut down their child processes, without separating them from the original system.
Wrong, jails have had the capability to cap cpu, RAM and other resources for some years now, as to clean child murder do the terms process group and session ring a bell?
 
Old 02-07-2014, 03:35 AM   #17
TobiSGD
Moderator
 
Registered: Dec 2009
Location: Germany
Distribution: Whatever fits the task best
Posts: 17,148
Blog Entries: 2

Rep: Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886
Quote:
Originally Posted by vl23 View Post
Wrong, jails have had the capability to cap cpu, RAM and other resources for some years now, as to clean child murder do the terms process group and session ring a bell?
Yes, jails have that capability, but cgroups does not have the other capabilities of jails, so saying cgroups = jails is plain wrong. If you want to compare jails with anything on Linux you might better compare it to LXC and/or chroot in conjunction with ggroups.
 
Old 02-07-2014, 06:13 AM   #18
vl23
Member
 
Registered: Mar 2009
Posts: 125

Original Poster
Rep: Reputation: 8
Quote:
Originally Posted by TobiSGD View Post
Yes, jails have that capability, but cgroups does not have the other capabilities of jails, so saying cgroups = jails is plain wrong. If you want to compare jails with anything on Linux you might better compare it to LXC and/or chroot in conjunction with ggroups.
So, to summarize:
Jails provide you with more features, and have been around longer, ergo they are a superior solution.

Also it seems like nobody is mentioning things like cpu sets, ulimits, quotas and the other UNIX/linux ways to cap resource usage that have been around since forever.
Again, to sum things up, cgroups is basically just a bunch of old tricks made easier to use by idiots and idiot applications with an ugly virtual file system interface that just clogs mount output.
 
1 members found this post helpful.
Old 02-07-2014, 06:43 AM   #19
TobiSGD
Moderator
 
Registered: Dec 2009
Location: Germany
Distribution: Whatever fits the task best
Posts: 17,148
Blog Entries: 2

Rep: Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886
Quote:
Originally Posted by vl23 View Post
So, to summarize:
Jails provide you with more features, and have been around longer, ergo they are a superior solution.
I wouldn't say that, because cgroups doesn't aim to have the functionality of jails. In the same way you could say that XFCE is a superior solution to LXDE, because it is around for longer and provides more functionality, but that would be missing the point in the same way.

Quote:
Also it seems like nobody is mentioning things like cpu sets, ulimits, quotas and the other UNIX/linux ways to cap resource usage that have been around since forever.
Again, to sum things up, cgroups is basically just a bunch of old tricks made easier to use by idiots and idiot applications with an ugly virtual file system interface that just clogs mount output.
If you want to see it that way, do it. I wouldn't consider that to be true. cgroups for example provides an easy way to track child processes and reliably kill them, something that is very hard to do using the existing tools. Making things easier is not always a bad thing.
 
Old 02-07-2014, 02:13 PM   #20
vl23
Member
 
Registered: Mar 2009
Posts: 125

Original Poster
Rep: Reputation: 8
Quote:
Originally Posted by TobiSGD View Post
If you want to see it that way, do it. I wouldn't consider that to be true. cgroups for example provides an easy way to track child processes and reliably kill them, something that is very hard to do using the existing tools. Making things easier is not always a bad thing.
Did you actually read any of my previous posts, hint hint, there is one about process groups above.
 
Old 02-18-2014, 07:00 AM   #21
Stuferus
Member
 
Registered: Jun 2013
Location: Germany
Distribution: Slackware
Posts: 174

Rep: Reputation: Disabled
i dont like changing what works so i vote and would vote in future for sysvinit!
also i did a test with a stopwatch.. my slackware boots faster than my mother mageia which uses systemd. 0:58 to 1:18.
 
Old 02-18-2014, 07:43 AM   #22
vl23
Member
 
Registered: Mar 2009
Posts: 125

Original Poster
Rep: Reputation: 8
Quote:
Originally Posted by Stuferus View Post
i dont like changing what works so i vote and would vote in future for sysvinit!
also i did a test with a stopwatch.. my slackware boots faster than my mother mageia which uses systemd. 0:58 to 1:18.
Lol, that is still slow compared to the test Gentoo on KVM I built, 11 seconds to boot into the equivalent of runlevel 3 then a few more (3-4 max)seconds to start X with twm.
Again.This is for a Virtual Machine!

Last edited by vl23; 02-18-2014 at 07:51 AM.
 
Old 02-18-2014, 09:13 AM   #23
TobiSGD
Moderator
 
Registered: Dec 2009
Location: Germany
Distribution: Whatever fits the task best
Posts: 17,148
Blog Entries: 2

Rep: Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886Reputation: 4886
Would be interesting to know which DEs are started, which services/daemons are started and so on. Just saying: "Hey, my Slackware boots faster than your Mandriva" or "Gentoo is faster than both" without sharing any details is as useless as benchmark as it can be.
 
Old 02-18-2014, 11:36 AM   #24
vl23
Member
 
Registered: Mar 2009
Posts: 125

Original Poster
Rep: Reputation: 8
Quote:
Originally Posted by TobiSGD View Post
Would be interesting to know which DEs are started, which services/daemons are started and so on. Just saying: "Hey, my Slackware boots faster than your Mandriva" or "Gentoo is faster than both" without sharing any details is as useless as benchmark as it can be.
ssh, ntp, an nginx web server and I am using dhcp on the system rather than static addresses(I should definitely disable that and go to static, since it will probably shave off a few parts of a second, considering how slow DHCP can be.)
All thad said I am planning to do a lot more where kernel compilation and other optimizations are concerned, for example I am only using -O2 compilation optimization since I am conservative, but i plan to try and use -O3 for a number of packages.
Still though I believe I am making decent headway for less than half a week of active experimentation/use.
 
Old 02-19-2014, 07:19 AM   #25
Stuferus
Member
 
Registered: Jun 2013
Location: Germany
Distribution: Slackware
Posts: 174

Rep: Reputation: Disabled
standard/default slackware64 14.1 to xfce without any autostart.

standard/default mageia 3 to kde4 without any autostart.

i did not change any package or service on installation or after. i installed both as is.

i tested my windows (i still have - shame on me!) too. 2:58, vista ultimate 32bit with autostart of antivir, fdm, drivers for mouse and gameing devices and programs displaying cpu and gpu temperature.

i think up to 1 minute is a acceptable time for an pc to boot, for me that feels fast enough!

vm = cheating! :P
 
Old 02-19-2014, 08:41 AM   #26
vl23
Member
 
Registered: Mar 2009
Posts: 125

Original Poster
Rep: Reputation: 8
Quote:
Originally Posted by Stuferus View Post

vm = cheating! :P
Um, why if anything VMs are slower than bare metal?
 
Old 02-19-2014, 11:02 AM   #27
jens
Senior Member
 
Registered: May 2004
Location: Belgium
Distribution: Debian, Slackware, Fedora
Posts: 1,463

Rep: Reputation: 299Reputation: 299Reputation: 299
Quote:
Originally Posted by vl23 View Post
Um, why if anything VMs are slower than bare metal?
Not for for boot time.
 
Old 02-19-2014, 11:12 AM   #28
vl23
Member
 
Registered: Mar 2009
Posts: 125

Original Poster
Rep: Reputation: 8
Quote:
Originally Posted by jens View Post
Not for for boot time.
LOL no, Bare metal:
appp-->kernel-->drivers--hardware
VM appinguest-->kernelinguest-->driversinguest-->virtualization infrastructure and host kernel virtualization components-->on-metal hardware
It would only be useful if you are using a jail, and that is not what I am doing, the I/O wait and memory access time would be higher since VM mapping would have to go through 2 kernels instead of one, and since the disk is a file on the host system.
 
Old 03-04-2014, 03:59 PM   #29
javaunixsolaris
Member
 
Registered: Jan 2007
Location: Colorado
Distribution: Kubuntu
Posts: 53

Rep: Reputation: 15
Quote:
Originally Posted by ReaperX7 View Post
If push comes to shove, I'll go back to Windows and use Cygwin.
Noooooooooooooooo!
 
Old 03-18-2014, 12:28 AM   #30
geox
Member
 
Registered: Jan 2012
Posts: 42

Rep: Reputation: 2
I like upstart a lot as it boots a lot faster than traditional SysV. I've written countless scripts for SysV and only started using upstart 6 months or so. Upstart scripts are a lot easier to write. Not just that, but I can easily specify when it should start (on filesystem, network, loopback interface, etc) and have all my boot processes start in parallel. And THAT is why SysV should be replaced.
 
  


Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is Off
HTML code is Off



Similar Threads
Thread Thread Starter Forum Replies Last Post
LXer: Debian May Be Leaning Towards Systemd Over Upstart LXer Syndicated Linux News 0 01-17-2014 03:30 PM
LXer: Debian Stil Debating Systemd vs. Upstart Init System LXer Syndicated Linux News 0 12-30-2013 06:02 PM
Debian To Replace SysVinit, Switch To Systemd Or Upstart jeremy Linux - News 0 10-28-2013 02:03 PM
[SOLVED] LPIC-1 updates: systemd and upstart matiasar Linux - Certification 11 09-25-2013 07:47 AM
Boot Delay 30min: systemd-analyze blame systemd-tmpfiles-setup.service BGHolmes Fedora 0 07-27-2011 09:02 AM

LinuxQuestions.org > Forums > Linux Forums > Linux - Distributions > Debian

All times are GMT -5. The time now is 03:55 AM.

Main Menu
Advertisement
My LQ
Write for LQ
LinuxQuestions.org is looking for people interested in writing Editorials, Articles, Reviews, and more. If you'd like to contribute content, let us know.
Main Menu
Syndicate
RSS1  Latest Threads
RSS1  LQ News
Twitter: @linuxquestions
Open Source Consulting | Domain Registration