On Thursday 13 November 2003 03:08 am, Cory Petkovsek wrote: : On Wed, Nov 12, 2003 at 09:38:46PM -0800, Ken Barber wrote: : > "Lazy sysadmins"? I beg to differ. : > : > How about "overworked sysadmins"? I was once in charge of a 'Doze : > network and there was no way I could keep current with the : > patches. Before one patch project was complete, there were two : > more vulnerabilities that needed patching. And patching M$ : > systems isn't exactly quick or easy with all of the testing that : > must be done first, not to mention trying to schedule the patch : > around various or department's schedules. I could have worked : > full-time at that place doing nothing else -- but I was required : > to do everything else. : > : > The problem isn't lazy sysadmins, unless not wanting to work 70 : > hours per week is your definition of "lazy." : : Agreed. In a recent thread about linux activism w/ numbers, I posted a : research report done by the Robert Frances Group, paid for by IBM. This : study focused on real world clients using windows, linux and solaris on : x86 and sun hardware respectively. They broke all measurements down to : normalized units so they could be compared fairly. They found from : these real world businesses that a windows administrator can generally : administer 10 servers, while a linux administrator, although more : expensive, can manage 44 servers.
>From my personal experience I would agree also. I must admit to using windows far less than linux. My experince with Service packs is that they often cause a whole lot of problems, which then have to be dealt with, which turns into a whole lot of time being eaten up to fix one problem that should never have been overlooked in the first place. Its a really frustrating problem that likely keeps admins from keeping windows machines current. >From the Real world experience files, I did my work co-op in a govt office. There were 170 workstations (at that time) running NT4, and we needed to update to SP4, IE, mcafee, and a few smaller apps. The bulk of the time was for IE, and NT4, and it took about 2 hours per workstation. The process was about 10 pages long, and the admin worked on one workstation at a time, at this rate, working by himself, he would have gotten it done just in time for the next update. If this were a linux shop, I could have taken my time (like all day If I wanted) to write a script to completely automate this process, and have all the workstations updated overnite (note, this means no downtime for the user too!). This way, I could have spent the rest of my time doing better things. I belive that the sql expliot (the nasty one earlier this year that took down bank of america, and many other large instituions, and servers all over the world) admins did know about well in advance, but did not install service packs. I think the blaster instance was well documented a couple years ago... I would have to look up references about it... It was a while ago, however I think it was TV news, not some crazy linux zelot that told me. : : http://www.rfgonline.com/subsforum/LinuxTCO.pdf : : Cory -- [In 'Doctor' mode], I spent a good ten minutes telling Emacs what I thought of it. (The response was, 'Perhaps you could try to be less abusive.') -- Matt Welsh _______________________________________________ EuG-LUG mailing list [EMAIL PROTECTED] http://mailman.efn.org/cgi-bin/listinfo/eug-lug
