Which distro has the smallest memory footprint on Linode?

Hi there,

Does anyone have any data points on this please?

Many thanks

Pete

8 Replies

Here's My Ubuntu 14.04 LTS Profile (Latest 64 bit (3.14.1-x86_64-linode39)), after fresh install and reboot:

top - 11:27:58 up 1 min, 1 user, load average: 0.02, 0.01, 0.01

Tasks: 72 total, 1 running, 71 sleeping, 0 stopped, 0 zombie

%Cpu(s): 0.0 us, 0.0 sy, 0.0 ni,100.0 id, 0.0 wa, 0.0 hi, 0.0 si, 0.0 st

KiB Mem: 2040900 total, 172508 used, 1868392 free, 3400 buffers

KiB Swap: 262140 total, 0 used, 262140 free. 78556 cached Mem

````
Distro #processes Mem usage/ MiB

Slackware 14.1 66 91
CentOS 6.5 72 97
Gentoo 66 100
Debian 7.5 70 111
openSUSE 13.1 74 113
Arch Linux 2014.04 73 115
Ubuntu 14.04 LTS 72 117
Fedora 20 72 139
````

  • Method: 'deploy a linux distribution' from Manager, then log in and run top

  • all on Latest 64 bit (3.14.1-x86_64-linode39) kernel

  • I did 3-4 reboots of each deployment, and there was some small variation in both observed variables

However, not really sure how meaningful or useful this info is (esp. in light of the 2Gb plans). I don't know what processes some distros spawn that others don't, or whether the difference is made up once application stacks are spun up and work is done.

With these numbers distro choice for smallest memory footprint (after boot) seems an irrelevant criterion.

Nice chart!

In other words, with a mere 48M spread from top to bottom (or 2.3% of the smallest Linode) it doesn't really matter which distro you chose.

How did you count?

I'm getting 29MB not counting what was accidentally cached at boot, on my Ubuntu 14.04 64bit Linodes.

Minimal installs though, but has the basic services running, cron, rsyslog, sshd, getty.

@vonskippy: that's certainly what it looks like after boot

@trippeh: I used 'top', which includes some buffers AFAIK. The figures are 10-20 Mb lower if I use 'free' -/+buffers/cache, so the difference between the distros may be entirely accounted for by the small differences in the processes they run up and how initial buffering is configured. In other words, there is no material difference between them :)

That said, I used the off-the-peg images available in the Linode Manager, and get around 80Mb for a clean Ubuntu 14.04 64bit with latest kernel, not counting buffers (mem command). Which image are you using, and how are you counting?

Once again, we see why there is quite simply never any valid reason whatsoever to ever use anything besides Slackware.

Buffers/cache can't be counted towards distro CPU use… For one thing, cache is available for applications if required.

>Buffers/cache can't be counted towards distro CPU use…

Do you mean memory use? I'd disagree though, take the case for example where the stock distro runs up processes and allocates buffers which are subsequently never used by applications? Buffer use is just as open to tuning as process choice.

Anyway, I looked a bit further what differentiates the different stock distros on Linode, and it is mostly in the choice of daemons for things such as dhcp, cron and logging, and whether they used systemd or not. It wasn't an exhausting look, but at least 1 distro (Debian 7.5) spun up some nfs related daemons as part of the stock configuration.

The conclusion of this short and incomplete trip around the distros is that there seems little difference in their memory footprint after boot :-) YMMV

Reply

Please enter an answer
Tips:

You can mention users to notify them: @username

You can use Markdown to format your question. For more examples see the Markdown Cheatsheet.

> I’m a blockquote.

I’m a blockquote.

[I'm a link] (https://www.google.com)

I'm a link

**I am bold** I am bold

*I am italicized* I am italicized

Community Code of Conduct