Fiber on LAN?
-
Hi,
I was wondering if I could get some tips on what im trying to accomplish on my test environment.So right now I have pfSense on hardware running on a ASRock H81M-VG4 R2.0 with Intel Celeron CPU G1820 @ 2.70GHz
2 CPUs. With TP-link using LAN 10/100/1000 which is connected to a dumb switch tp link then to a test computer nothing especial.I want to upgrade to test it out with fiber. Im planning to buy a Ubiquiti-EdgeSwitch-ES-8-150W-Gigabit-8-Ports-24-48V-POE-Fully-Managed-Switch which has SPF, I was thinking to also buy a PCIe fiber NIC to replace the TP-link and test the speeds.
My question is which NICs are best supported for pfSense 2.2.6 and is it worth running fiber on a production area around 50-100 computers, using also SQL server.
The idea is I would run test before actually presenting the results to my boss, as of right now we have 2 HP switch Procurve 2920-46G without using the SPF ports and a dell Rack R720(with the SQL machine virtualized). As there has been complaints about speed on the LAN so I was thinking on changing the infrastructure
Thank you
-
The only reason to run gig fiber over copper is to increase the distance between switches. You can run 550m over multi-mode and like 10km+ single mode. Even farther with LR transceivers. Copper's limited to 100m. Fiber is not any faster or more reliable than copper.
There might also be some benefit in getting multiple strands/channels over short distances in smaller cable diameters. You can get a 24-strand (12 full-duplex paths) in the same diameter as about 3 cat-5 runs.
For gig, Intel copper NICs are still your best bet. If you need to haul that over fiber, use your SFP ports in your switch.
-
Fiber is not any faster or more reliable than copper.
Let's see… copper + lightning strike = BOOM. Bye bye switches!
One benefit of fiber is that you don't have to use surge protection, as they come with one built-in. -
That's another one. Thanks. It's still not any faster.
-
Well…I dont even know what to do been getting a 2nd opinion on the Lag that supposedly its the network which i dont see it and nor do they. Which now i think it might not be the network rather then SQL database, But I am Curious How the speeds would increase if connecting a patch cord of fiber to switch1 to switch 2 instead of connecting Cat6
-
It wouldn't unless by doing so you were eliminating a source of errors or something.
If you're looking at correcting an application performance issue by going from gig copper to gig fiber you are wasting your time.
You can use wireshark to zero in on TCP connections and look at the packet delays. You might want to do that on the SQL connections instead.
-
This post is deleted! -
-