Best GB switch for the money?

Dec 5, 2005
247
0
0
I'm moving into a house with a friend of mine, and we're going to pull some bundled media cable with coax/ethernet to each room. In the central location for the wiring we're going to put in a gigabit switch. I'm thinking about just getting a netgear 8 port gigabit switch for around $70, Probably this guy. Would anyone suggest a different gigabit switch for 8-12 ports???
 

bluestrobe

Platinum Member
Aug 15, 2004
2,033
1
0
I spend more time replacing netgear stuff than using it. I got the Dell Power Connect 2708 new off of eBay for $50.00 shipped. If you wish it even has the option of being a managed switch.
 

Madwand1

Diamond Member
Jan 23, 2006
3,309
0
76
Jumbo frames are somewhat overrated IMO, but the 2708 doesn't support them.

Isn't the 2724 actively cooled? So couldn't it be noisy at home? Then a 2716 might be a better match midway.

But please tell me, what makes the Dells better than consumer Netgears for example in a home setting?

Here's a sample of my GS608's performance capability (with 9K jumbo frames). Not to be taken as any sort of real-world test, but as one demonstration of the switch's capability.

F:\tools\bench>iperf\iperf -c 192.168.0.135 -l 256k -t 21 -i 3 -r
------------------------------------------------------------
Server listening on TCP port 5001
TCP window size: 8.00 KByte (default)
------------------------------------------------------------
------------------------------------------------------------
Client connecting to 192.168.0.135, TCP port 5001
TCP window size: 8.00 KByte (default)
------------------------------------------------------------
[720] local 192.168.0.100 port 8622 connected with 192.168.0.135 port 5001
[ ID] Interval Transfer Bandwidth
[720] 0.0- 3.0 sec 85.5 MBytes 239 Mbits/sec
[720] 3.0- 6.0 sec 354 MBytes 990 Mbits/sec
[720] 6.0- 9.0 sec 355 MBytes 991 Mbits/sec
[720] 9.0-12.0 sec 354 MBytes 990 Mbits/sec
[720] 12.0-15.0 sec 355 MBytes 991 Mbits/sec
[720] 15.0-18.0 sec 354 MBytes 990 Mbits/sec
[720] 18.0-21.0 sec 354 MBytes 989 Mbits/sec
[720] 0.0-21.0 sec 2.16 GBytes 882 Mbits/sec
[700] local 192.168.0.100 port 5001 connected with 192.168.0.135 port 1426
[ ID] Interval Transfer Bandwidth
[700] 0.0- 3.0 sec 355 MBytes 993 Mbits/sec
[700] 3.0- 6.0 sec 354 MBytes 988 Mbits/sec
[700] 6.0- 9.0 sec 354 MBytes 991 Mbits/sec
[700] 9.0-12.0 sec 354 MBytes 989 Mbits/sec
[700] 12.0-15.0 sec 354 MBytes 990 Mbits/sec
[700] 15.0-18.0 sec 354 MBytes 991 Mbits/sec
[700] 18.0-21.0 sec 354 MBytes 988 Mbits/sec
[700] 0.0-21.0 sec 2.42 GBytes 990 Mbits/sec

Additional commentary. The Netgear is better than a D-Link that I have on hand (DGS-1008D rev C3) -- which is fine for standard frames, but its performance decreases when jumbo frames are enabled, pretty much defeating the point of having jumbo frames.
 

JackMDS

Elite Member
Super Moderator
Oct 25, 1999
29,534
416
126
It issue with Jumbo Frames boils down to the topology.

If you configure a separate Giga network that does not connect to the Internet it better to shop a littler more and get Jumbo frames.

If the Giga goes together with your Internet activities, there is no point for the Jumbo since you are not going to configure a TCP/IP that goes to the Internet with 9000 MTU.

Peer to Peer Giga Networks

Setting a Home/SOHO Giga network.

:sun:
 

Madwand1

Diamond Member
Jan 23, 2006
3,309
0
76
I don't run separate networks or subnets here, and don't have any problems of note mixing and matching jumbo frames -- I guess path MTU detection is doing its job. The only "internet" problem I have is that I can't get to my (non-JF capable, GbE) router's web config page when JF are enabled, but there's a simple fix here -- click on the web page 3 more times, wait for timeouts, and then remember "oh yeah" (j/k) and disable JF on my NIC or change to a computer that doesn't support JFs.

The problems I have with JFs are more subtle -- spending time and energy in obtaining devices that actually support JFs at decent sizes, dealing with the occassional performance loss accompanying jumbo frames, and often finding after this effort that the returns are marginal.

Returns are significant when I'm dealing with PCI NICs at high bandwidth, or with high CPU utilization, thus it's worth the hassel in some cases.

But I think I have to agree in general that the potential minuses aren't worth the potential plusses in many cases, hence the comment about JFs being overrated.