Industry Standard?

Corporations/businesses usually are setup with a switch or two on each floor (depending on the floor and the quantity of people in each floor). Almost always a whole switch does not go down -- it's usually one or two ports that fail and most of the time it's because of a vaccuum cleaner eating a cable or the cable being jolted out of the socket. And usually you can replace the cable or switch the person onto another switch socket while you diagnose/troubleshoot.

Governments are setup similarly as distance permits. Ethernet has a distance limitation and switches extand that limitation. You ALWAYS want a few extra (known-working) switches to be setup and possibly placed into the loop should one of them fail. Depending on distance (and if you are in an "L" shaped building) you can either place a central switch in the middle of the "L", or one at each end. You generally do not want outages scattered throughout the building.... you want the problem to be in one location. It's easier to find problems this way.

I am just gonna post my thoughs on this one. Others will probably chime in with URLs to documents you can read on network setup. But here is what I think on your questions.
It would be nice to put each department on its own switch or VLAN in a large switch (like a Cisco Catalyst 6509). Put routers between the switches so each switch can be a different IP subnetted network (192.168.1.0/24, 192.168.2.0/24, etc). You keep your departments on different wings in the same building. So, a different switch per wing (per dept), and probably a backup switch per wing also set up, so that if one switch goes down, you can fail over to the 2nd switch.
The routers are also set up to have a backup router for each one, again in case of disaster. YOu could run HSRP on the router pairs for this redundancy.
This would be nice. So, if a router goes down, the 2nd router in the pair would automatically pick up, keeping the traffic flowning. If a switch goes down, you migrate to the backup switch. Trafficcontinues to flow.
You also should use a star-shaped configuration, so that no department/wing is logically located BEHIND another department/wing. This is so a total outage in one dept does not sever all connections to ANOTHER dept.
Let's see, you would keep all switches in their pairs on different AC currents and on different UPS devices; same goes for the routers. If you lost power on one circuit, then you would only lose one switch and/or 1 router; the other set would be on a different circuit and would stay up.

I guess I should have been more specific. Right now the budget is on hold due to the change in administration this past winter. We just purchased two switches with 4-5 blades each, and each blade having 38 ports (I figured in the few blades that didn't have as many). Which is a total of about 345 ports. We just installed another blade that is not being used; we were anticipating the re-configuration of this network. There are approximately 250+ computers that are connected to this network. As I stated before, the budget is nonexistent so we do not have any room to purchase anything new, in fact we have to go as far as reuse "store-bought" cables and crimp them down to size (right now they are 10' when they only need to be 4-5! Just imagine that rat?s nest of wires!). We are running 6 servers off of this network, some of them being for printers, others for our website, backups, etc. I am assuming from my research the best thing to do would be split evenly distribute the serversbetween the switches, not placing two servers on a single blade. Not only do we have different departments on each floor, we have six conference rooms on each floor. However, the idiots that designed the layout did not put the administration foreach department on that same floor. We have 8 departments, with a few other groupings of people thrown in there. I was thinking of putting each department on their own blade, except in one case where there are more computers than ports, and even though it's not on the same floor, keeping the administration with their respective department, and go from there in terms of the few extra groups of employees.

You may want to speak to your Business Continuity / Contingency / Disaster Recovery group. Being in this business, I've found that they can help out quite a bit, on the topology design and funding of redundency.

This question was auto-closed because it is 90 days old