System Reboot

Nathaniel Hall halln at otc.edu
Wed Dec 8 05:01:04 UTC 2004


The other way to think about rebooting a machine can be from a security 
standpoint.  If the machine has not been rebooted, then it hasn't had 
its kernel updated.

Nathaniel Hall, GSEC
Intrusion Detection and Firewall Technician
Ozarks Technical Community College -- Office of Computer Networking

halln at otc.edu
417-447-7535



Dave Ihnat wrote:

>On Tue, Dec 07, 2004 at 04:57:45PM -0600, Rich Ransom wrote:
>  
>
>> 17:03:45  up 458 days,  9:46,  1 user,  load average: 0.02, 0.07, 0.05
>>
>>Its running our office mail server, our company website with php and 
>>postgres. Everything works fine so is there any reason to do a system 
>>restart?
>>    
>>
>
>Heh.  There are two schools of thought on this.  One figures, "It
>ain't broke, don't fix it", plus they figure that long uptimes give
>'em bragging rights over Win boxen.
>
>The other--and I happen to live in this one--believes that, while we don't
>*need* to reboot on a regular basis (unlike every Windows box I've ever
>managed--not a snark, just stark fact), if something IS broken, it usually
>shows up in a reboot.  It's better to do a reboot periodically when YOU
>can afford to find something is broken, than have it happen unexpectedly.
>
>Plus, every complex system has bugs, and far too many C and C++ programs
>have memory leaks.  Even if they're small enough that you can run for
>over a year, it can't *hurt* to periodically reboot and clean things up.
>
>So unless you're big on bragging rights, I'd schedule a reboot and system
>check every 90 days or so.
>
>Cheers,
>  
>



More information about the redhat-list mailing list