I modified a Cisco 3560 to be my home network switch.

Cardboard

PRE-EDIT – I typed this the first time in neatly separated paragraphs. We will now see if this forum is still clumping things all together. I typed it in a Word document and Copy and Pasted it into the text box here. Short version – I modified the fans. Long version – My wife decided she wanted to save a few bucks by switching us from Comcast to Verizon FIOS. I have my network set up with the modem in another room, and an ethernet cable runs to my room, to a Netgear WGT624 v4 4 port switch / wireless combo, then out to my PC and others. Before I did anything, I ran speed tests to several cities. I was getting 42 down and 6 up. While the FIOS guy was hooking everything up, he said something about Comcast being able to detect speed tests running, and throttling my neighbors’ bandwidth to give me better scores. I don’t know about that. Once it was connected, I ran the speed tests again. 45 down and 38 up. I was supposed to be getting 150/150. I hooked the modems’ ethernet cable directly to my PC, and got 150 down and 160 up. Seems like that old Netgear box is a slight restriction. So I hooked up one of my WS C3560 24 PS-S 24 port switches from my CCNA live test equipment. I put the modem on port 1 and my pc on port 24. It scored 94 down and 98 up. Not perfect, but much better. Cisco specs say it has 24 10/100 ethernet ports and 2 Gigabit SFPs, so is this about the top end then, for its speed potential, to speed test at 94-98 Mbps? I think I’d need a 3560G switch to get better speeds. I saw on the Cisco website that there were about 4 different types of SFP transceivers. If I wanted one to plug an ethernet cable into, which would be best for my needs? So the only problem with this switch is the fans. Way too loud! I took the cover off with it running and looked at everything. It you unplug one of the 3 fans, the others go to top speed. So I disconnected all 3 stock Cisco fans, and I pulled a standard 80mm fan out of an old desktop, and after rearranging the 3 power leads in its plug to match the Cisco fans, and testing with a multimeter, it is in place and even though it runs at its top speed, it is far quieter. I removed the fan from the side of the case so I could run the power leads through the hole, and I taped the fan to the outside of the switch (the switch is standing on end under the table). The 80mm fan blows air in through the power supply, and it has been fine for 2 weeks and 3 days so far. So it has been a successful project.

Cardboard

Yep, all my carefully typed paragraphs are clumped together. You'd think that on a tech computer forum, the person in charge would be able to handle an issue like this...... :rolleyes:

Priston

Netgear WGT624 is 10 years old and has only 10/100 ports.
WS-C3560-24-PS is even older and has only 10/100 ports.

There are way more than 4 different kinds of SFPs. Assuming the ethernet cable you're talking about is twisted pair. GLC-T and SFP-GE-T are your choices. SFP-GE-T has a temperature sensor. GLC-T does not.

Instead of spending money on SFPs or an old switch, why not simply buy a new home wireless router?

Cardboard

If the Netgear and Cisco switches are both 10/100, why was the Netgear so much slower? I tested them both to multiple cities (the same cities), and the results were consistent. Ethernet cable in question is normal Cat 5E.

Priston

I wish I had a good website that explained packet routing throughput and frame switching throughput.

While the Netgear router has 100mbps ports, the maximum throughput could be less than 100mbps on the layer 3 side.