[CentOS] reboot long uptimes?

Tue Feb 13 15:03:52 UTC 2007
Drew Weaver <drew.weaver at thenap.com>

 

-----Original Message-----
From: centos-bounces at centos.org [mailto:centos-bounces at centos.org] On
Behalf Of Johnny Hughes
Sent: Tuesday, February 13, 2007 6:30 AM
To: CentOS ML
Subject: Re: [CentOS] reboot long uptimes?

On Tue, 2007-02-13 at 12:06 +0100, D Ivago wrote:
> Hi,
> 
> I was just wondering if I should reboot some servers that are running 
> over 180 days?
> 
> They are still stable and have no problems, also top shows no zombie 
> processes or  such, but maybe it's better for the hardware (like ext3 
> disk checks f.e.) to reboot  every six months...

I only reboot on kernel upgrades, that is usually more often than 6
months.  But if you don't need to reboot for that reason, I would not
reboot at all.

> 
> btw  this uptime really confirms me how stable Centos 4.x really is 
> and so  I wonder how long some people's uptimes on the list are ;)
> 
> rmc

You should consider upgrading your kernels when security updates come
out ... just to be safe.  Especially for machines touching the internet.

I usually upgrade my kernels because I like to use LVM snapshots for
backups and that has only really started working semi-well since 4.3 and
even better in 4.4 ... so most of my machines get rebooted every new
kernel, which is at least 2-3 times a year (sometimes more often).

That being said, I do have a non internet facing machine that has not
been rebooted since it was installed with CentOS-4.0 on it one March 1,
2005.  It is an internal router on my employer's infrastructure, and has
been up for almost 2 years (and was installed on the day before CentOS-4
was officially released).

Thanks,
Johnny Hughes
-------------

	My uptime on some of our boxes are pretty bad, we have roughly
250 CentOS 4.x boxes here I'd say probably 25% of them initially suffer
from some sort of bug with cpuspeed which causes kernel panics (until we
disable cpuspeed), and then we have this other curious thing that
happens with the filesystem where they will occasionally start spamming
this "ext3-fs "Journal Has aborted" message until we reboot the boxes
(nothing is wrong with the hardware in any of the cases).

Other than those 75 or so issues no problems at all.

-Drew