[Gllug] Linux in a Production Environment

The Flying Hamster hamster at vom.tm
Fri Jun 22 12:00:05 UTC 2001


On Thu, Jun 14, 2001 at 09:09:23AM +0100, Gordon Joly wrote:
> >
> >I use it as my organisations main file server, print server, web server,
> >mail server, dns server, fax server, and remate access server. Max
> >uptime on my main machine was just over 500 days. It needs *far* less
> >attention than my NT boxes.
> 
> 500 days uptime is of course better than most systems, but how do you 
> maintain it without downtimes?

Most maintenance can be done (if planned properly) without taking down
the machine or the service (apache for example can be reconfigured
without interupting service and restarted in the space of a second or
two).

> For example, security patches and upgrades to the kernel?

Depends on the nature of the security hole, some security holes
require shell or user access on the server, these will only require a
reboot if there are users on the system.  Certainly all but one or two
machines here don't have the users anywhere near them at a shell level.
 
> Also, other patches that require (or suggest) a reboot?

Not come across one of these on Linux.

NT/other M$ware... well sneeze near it and it prompts for a reboot.
>From personal experience of trying to manage people and time here is
that m$ware should be avoided.  I know that a single admin or two can
keep this network running, that's all the Linux boxen, switches,
routers and wan links.  To keep the exchange and IIS boxen running
(total = 3) I need the same number of people and they're constantly
overloaded. 


-- 
The Flying Hamster <hamster at suespammers.org>         http://hamster.wibble.org/
Discoveries are made by not following instructions.

-- 
Gllug mailing list  -  Gllug at linux.co.uk
http://list.ftech.net/mailman/listinfo/gllug




More information about the GLLUG mailing list