On Sun, 2002-06-02 at 20:48, Jeff Waugh wrote: > <quote who="David Kempe"> > > > My experience is that I have 50 user, 3 server clients that have no > > fulltime support at all. In fact - for the scenarios I am thinking of the > > ratios would be mostly similar - sorry to burst your bubble on this. > > That's the critical difference though - you're comparing 50 user, casually > maintained, low policy (security, standards) sites to far more demanding, > multiple location, high policy, greater than 200 user sites. Internal IT > departments and support, remote user support (ie. laptops), etc. It's a > different world.
The other thing here is that in smaller orgs, there is usually an internal contact (support person) or two who handle the niggly little problems, they only pass on the bigger problems to an outside party. For larger orgs, they expect to cope with all the niggles and thus must account for them when selecting staff for IT support. Thus in your cases, you'd need to count 2 people - niggle fixer and your support person - thats 1:25. I was wondering what experience has found is the ratio of those bigger sites - Linux and MS? Some of my own annecdotal (wish I could spel) - A North Shore Council that is a Windows shop I happen to know has 8 IT staff for 300 users (1:37.5 users). They do little or no internal dev, thats just supporters and maintainers and niggle fixers. - I know another windows Council that has 40 staff for 500 users (1:12.5). - I was talking to a guy from a local Uni, they are running more than 500 Linux clients NFS. I forgot to ask him about the support staff but I will. - Couple of Linux Installs similar to your story. Biggest 50 users virtually no support required but an internal guy who handles the niggles. - Solaris/Windows install - samba for LAN access. 4 staff IT handle all 300 users BUT every state (6 - NZ is the Eastern State) has 1 person who handles all support calls first before forwarding if unhandleable to the IT guys/gals (I calc this as 10:300 - 1:30). These are the sort of ratios I'm looking for but hopefully in a more scientific study I can point to and say - look here.. etc etc. > > > A well setup windows site with competent admin should probably have about > > the same with a well setup linux site. Sure it will almost definitely cost > > more to get to that state, but the TCO running costs should be similar. > > Keep also in mind that Linux based network are not immune to be set up > > badly. I don't agree. I think it is currently much more difficult to administer, run and control a pure windows shop. I think it costs more money than a Linux shop for day to day IT support wages. That's my experience - can anyone show me a study of such. > > TCO calculations generally assume a reasonable setup and competent > administration (otherwise the data would be close to useless), but may > include disaster recovery components, etc. True but they rarely include how many bodies in the IT shop. > Unfortunately for us, the Free Software choice is a long term proposition. > We have to compete with and/or deliver on short term business goals to > succeed. We could learn a few tricks from NSW salination protestors and > working groups, and some centre-left environment groups on this point. > That's why I need a 6 year TCO that covers staffing, licensing, support and other factors not mentioned by MS as they are a not useful to their argument. Of course we blow them out of the water on license fees but that is not the whole picture. We all know that once installed, most Linux servers are 'set and forget' surely there are studies to prove it. -- SLUG - Sydney Linux User's Group - http://slug.org.au/ More Info: http://lists.slug.org.au/listinfo/slug
