Reliability Survey: Windows Servers Beat Linux Boxes

Yahoo News ^ | Mon Jun 5 | Gregg Keizer

[zeugma]

Well, your load average is higher :-)

The two 1300+ day systems I have are internal webservers that are a bit overpowered for their current load. A lot of their processes have been offloaded to newer systems. They are scheduled to be decomissioned in the next 6 months. 

[antiRepublicrat]

Sure you do and Ill tell you why..

That's not the way I meant it. I meant people care about how many hours per year is this web site available. Both the OS and the service affect that. IOW, I meant the OS is only part of the picture, and Microsoft has some things beyond the base OS that make it look even worse.

[zeugma]

In addition to the 'uptime' command, you can also use 'last'.

 bash-2.05# uptime
 10:14am  up 236 day(s), 17:02,  1 user,  load average: 0.28, 0.17, 0.19
bash-2.05# last reboot
reboot    system boot                   Thu Oct 13 17:12
reboot    system boot                   Thu Oct 13 17:04
reboot    system boot                   Tue Feb 15 21:15
reboot    system boot                   Tue Feb 15 21:04
reboot    system boot                   Sun Nov 14 19:50
reboot    system boot                   Mon Oct 18 10:16
reboot    system boot                   Sat Oct 16 03:55

wtmp begins Sat Oct 16 03:55
bash-2.05#

Unfortunately, the 'last' command is of not much use if you measure uptime in years...

$ uptime
 10:16am  up 1349 day(s), 17:16,  1 user,  load average: 0.01, 0.01, 0.01
fada1w3p11: pugha :) last reboot
reboot    system boot                   Thu Sep 26 17:00
reboot    system boot                   Sun Aug 11 16:34
reboot    system boot                   Sun Aug  4 15:33
reboot    system boot                   Sun Aug  4 15:30
reboot    system boot                   Sun Aug  4 15:27
reboot    system boot                   Fri Aug  2 15:35
reboot    system boot                   Fri Aug  2 15:27
reboot    system boot                   Fri Jul 26 08:01
reboot    system boot                   Thu Jul 25 17:05

wtmp begins Thu Jul 25 17:05
$

So, the system was rebooted sept 26. But it doesn't tell you that was Sept 6., 2003!

[N3WBI3]

Apologies and good point

[N3WBI3]

If youre using log rotate last is not going to be too much help..

[zeugma]

Yup. Logrotate kills the usefulness of last. You should see the size of the wtmp files we have on some of our systems.

[N3WBI3]

Given it catcehs you last reboot I cant even Imagine how big it is..

The system I put up gets a bit of traffic so I cant leave logs out there or I would fill up the drive..

[zeugma]

I just took a look at it, and it doesn't look as bad as I'd have thought. wtmpx is about 7MB.

The system is running Solaris 8 and is set up to only log at 'error' or above.

I've had multi-hundred meg wtmp files on some systems, so I really prefer to set logrotate to catch them before they get out of hand.

[N3WBI3]

Good article debunking a previous study on the same subject by the same group..

http://www.businessweek.com/the_thread/techbeat/archives/2005/04/the_truth_about_1.html

[amigatec]

What about the Novell Server that had over 6 years of Uptime??

[amigatec]

My Linux firewall:

16:10:53 up 262 days, 8:39, 1 user, load average: 1.84, 1.38, 1.26

[rzeznikj at stout]

You both have me beat...8^)

4: uptime
16:56 up 318 days, 12:01, 40 users, load average: 0.53, 0.53, 0.53

[Halfmanhalfamazing]

If you search my posts you'll see that this is probably the first time I've brought that up.....

Well, I didn't explicitly say it, but I clearly implied it.

Most of the time, their claims can be thrown out on their face, no need to play the silly discreditism game.

[Halfmanhalfamazing]

suse10.1@linux-49in:~> uptime

10:47pm up 1:16, 3 users, load average: 0.07, 0.05, 0.07

:-)

[rzeznikj at stout]

Mine was when I was connected to the Unix server at school...

This is for my Suse 10.0 box...

9:48pm up 1:08, 3 users, load average: 0.72, 0.67, 0.48

[Halfmanhalfamazing]

Ah hah! AH HAH!!! SEE!! I BEAT YOU BY 8 MINUTES!

lol

:-P