will a switch slow my conection down?

ic0n

ic0n

ic0n

Associate
Joined
12 Oct 2005
Posts
850
Location
left abit no your other l
hi guys

i have a nice fast connection some days (come from a university) tested to day and came in at 72.47 Mbps i have had it running faster over 100Mbps before so would a switch limit my speed?
 
no

a switch doesn't split a connection, but obv if you've got 3 pc's connected to it and they're all trying to download from that line, they won't get 72 each
 
ok cool thanks so it wont limit the speed in any way as long as theres nothing else using the connection?
 
yeah

if you've got say 5 pc's connected to the switch, and they're just in windows doing nothing, you won't notice at all

your pings will rise by a small ammount, usually 1-15ms, depending on the cable length/quality

^not noticable
 
Using a switch is recommended, if you reset your router your LAN I/O is not interrupted (except for on line access at that time) ie I can copy files from my HD to the NAS, switch router off and it just continues.
 
You connect a single RJ 45 from your router to the switch, and all computers plug into the switch. Make sure you buy a switch with enough connectors, you soon fill them up. I have a 8 port 100 Negear, using 6.
 
bledd. said:
yeah

if you've got say 5 pc's connected to the switch, and they're just in windows doing nothing, you won't notice at all

your pings will rise by a small ammount, usually 1-15ms, depending on the cable length/quality

^not noticable
Click to expand...

Your pings won't rise by 1-15msec. Unless you've got a dodgy/faulty switch.
 
oh right, i always assumed they'd rise a bit :) -mainly due to cable length/quality

the only long cable i've used before was about 35m cat5 (not e) and was walked all over :D, added about 10ms to our CS lan pings
 
bledd. said:
oh right, i always assumed they'd rise a bit :) -mainly due to cable length/quality

the only long cable i've used before was about 35m cat5 (not e) and was walked all over :D, added about 10ms to our CS lan pings
Click to expand...

Probably due to it being knackered, yeah :p

I once thought that too but the actual switching speed is usually in the nano second region so the difference in ping is close to zero (assuming the cable is working correctly). I've ran cat5e at pretty much the limits of it's spec and even compared to a cat6 2 meter cable, the PC at the end of the cat5e 100 meter cable had very similar pings :) Similar enough to not be able to measure the difference anyhow.
 
bledd. said:
oh right, i always assumed they'd rise a bit :) -mainly due to cable length/quality

the only long cable i've used before was about 35m cat5 (not e) and was walked all over :D, added about 10ms to our CS lan pings
Click to expand...

Highly unlikely I think...you have to remember that a ping is not an accurate measurement of your network. It's only designed to test connectivity between devices. I've been to places where we've had sub 10ms pings from Wokingham to Epsom yet none of the devices could register to the servers in Epsom due to poor quality of service (QoS). Took hours to convince the service provider to look into the issue properly and low and behold they'd forgotten to configure the QoS. Once done, we still had sub 10ms pings but the devices could then register with the servers.

There's a whole science behind configuring networks and a network engineer worth his/her salt wouldn't really on pings to test the quality of a network ;)
 
Yeah ping is a basic tool for checking connectivity, it is in no way a decent method of measuring performance.
 
You must log in or register to reply here.
Back
Top Bottom