Technical Discussion
  >> Home Networking, Internet Connection Sharing, etc.


Register (or login) on our website and you will not see this ad.


Pages in this thread: 1 | 2 | 3 | 4 | 5 | >> (show all)   Print Thread
Standard User joconnell
(experienced) Mon 30-May-11 20:57:36
Print Post

Measuring home network speed


[link to this post]
 
I never realised this could be so hit and miss!

I have a cat5e wired network at home and decided to try to determine the true network speed. I tried Ixia QCheck but this reported network speeds of 200Mbps but didn't work consistently and would crash, so I stopped using it.

Next up, I downloaded iPerf and tried that, but it reported speeds of 9Mbps.

Thinking something is seriously wrong, I plugged my two PC's NICs directly into my Netgear 5 port gigabit switch (it's a GS605) using 2m cat5e patch leads and re-ran the tests - little difference when testing with iPerf.

I decided to try some other testing utility so I downloaded LAN Speed Test and tried that using the LST client and LST Server.

Using LST version 1.x, I would occasionally see speeds of around 800Mbps using a 500MB test file, bit usually the speed was reported at around 230Mbps.

I tried LST version 2 which allows you to set the number of packets and packet size for the network write (to server) and read (from server) and got a reported speed of around 240Mbps as you can see below.

Text
1
23
45
67
89
1011
1213
1415
1617
1819
2021
2223
2425
2627
Packet Length      Write Speed        Read Speed
-----------------  -----------------  -----------------52,428,800  100   164.0118408 Mbps   294.8089371 Mbps
62,914,560  85    161.1970444 Mbps   295.0499115 Mbps5,242,880  10     154.5328293 Mbps   288.1078339 Mbps
5,242,880  10     158.5221329 Mbps   291.6319733 Mbps5,242,880  50     142.5855789 Mbps   282.5011215 Mbps
5,242,880  15     156.6718521 Mbps   288.5421829 Mbps1,048,576  25     159.1758652 Mbps   271.3827896 Mbps
1,048,576  25     159.4061890 Mbps   275.7951202 Mbps1,048,576  100    159.1510773 Mbps   274.8109436 Mbps
1,048,576  10     157.4700623 Mbps   277.1000748 Mbps1,048,576  10     157.8203354 Mbps   275.6318588 Mbps
1,048,576  10     160.0641785 Mbps   274.6975479 Mbps1,048,576  100    159.8181458 Mbps   274.7813950 Mbps
1,048,576  100    160.0971909 Mbps   269.0993423 Mbps10,485,760  10    160.9982986 Mbps   243.6563797 Mbps
10,485,760  100   163.2350769 Mbps   243.2745895 Mbps1,048,576  10     157.0954742 Mbps   211.3537216 Mbps
1,048,576  10     158.8932648 Mbps   216.5071106 Mbps1,048,576  10     160.9317627 Mbps   222.2300491 Mbps
1,048,576  10     160.9545670 Mbps   217.3642349 Mbps52,428,800  10    159.2238083 Mbps   236.5655975 Mbps
52,428,800  10    161.2564011 Mbps   236.9036484 Mbps10,485,760  10    162.1253204 Mbps   236.3230057 Mbps
10,485,760  50    163.0046539 Mbps   240.9702911 Mbps10,485,760  50    160.0156631 Mbps   237.9985428 Mbps


Remember, this is the network consisting of 2 PCs with gigabit NICs, connected to a gigabit switch with 2m cat5e patch leads, nothing else in the equation, so I'd expect performance to be a lot better than above.

I'm not sure if testing my LAN accurately is a possibility, but I guess that if I test the actual home LAN and get figures comparable to the above, then it's performance will be as good as I can get to my test LAN as described.

Has anyone else here tried testing their wired ethernet LAN? Any advice or recommendations?

My Broadband Speed Test
Standard User MHC
(legend) Mon 30-May-11 21:35:19
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
Do your NIC have a test facility? My desktop PC has NIC diagnostics and that tests the cable and link. It reports back 1Gb.

Not perfectly accurate but a good estimate is to o a PC to PC transfer of very large files - I use a set of images at around 1GB in size. That should take 8 seconds but was between 10 and 11 giving a speed of 700+ Mbps which is around the best the hard drives can sustain.





~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


M H C


taurus excreta cerebrum vincit
Anonymous
(Unregistered)Mon 30-May-11 21:55:26
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
Could the read and write speeds be limited by the hard drive speeds, and not the network speeds.
Tony


Register (or login) on our website and you will not see this ad.

Standard User awoodland
(regular) Mon 30-May-11 22:14:38
Print Post

Re: Measuring home network speed


[re: Anonymous] [link to this post]
 
Could also quite possibly end up being CPU-bound depending on the hardware.
Standard User joconnell
(experienced) Mon 30-May-11 22:27:03
Print Post

Re: Measuring home network speed


[re: MHC] [link to this post]
 
In reply to a post by MHC:
Do your NIC have a test facility? My desktop PC has NIC diagnostics and that tests the cable and link. It reports back 1Gb.

Not perfectly accurate but a good estimate is to o a PC to PC transfer of very large files - I use a set of images at around 1GB in size. That should take 8 seconds but was between 10 and 11 giving a speed of 700+ Mbps which is around the best the hard drives can sustain.

The network status dialog shows the connection as 1GB but that's probably not based on any test.

I copied 502 files totalling 1.32GB to the other PC's shared folder and it took 4mins 54 secs which is way way slower than your test result. A 1.94GB video file took 6.5 minutes, which seems very slow, but plays back over the network without any initial loading delay and I can navigate around the video as if it were a local file.

I'm wondering if the fact that the second PC has 2 network cards would impact perfromance, or that the second PC is quite old (circa 2001)?

My Broadband Speed Test
Standard User joconnell
(experienced) Mon 30-May-11 22:28:39
Print Post

Re: Measuring home network speed


[re: Anonymous] [link to this post]
 
In reply to a post by Anonymous:
Could the read and write speeds be limited by the hard drive speeds, and not the network speeds.
Tony
The speed test writes and reads to a remote server process which doesn't create any file but just uses memory, so disk speeds shouldn't be an issue.

My Broadband Speed Test
Standard User MHC
(legend) Mon 30-May-11 22:43:06
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
Network Status means very little ... Is there not a NIC set up application with test capability?

Your results are around 35-40 Mb ... way down. Dual NIC should not have any impact, unless you are trying to read and write to the same disc from both NICs.

Use TBB meter or Windows Task manager to get a display of traffic and do another transfer. Watch the trace - is it stable/regular, does it peak and fall back? It will give you another indication of speed.

Have you checked the cables? Is there the possibility that there is a nasty noise source causing problems on the network? Or maybe a NIC is failing?





~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


M H C


taurus excreta cerebrum vincit
Standard User joconnell
(experienced) Mon 30-May-11 23:41:54
Print Post

Re: Measuring home network speed


[re: MHC] [link to this post]
 
In reply to a post by MHC:
Use TBB meter or Windows Task manager to get a display of traffic and do another transfer. Watch the trace - is it stable/regular, does it peak and fall back? It will give you another indication of speed.

Have you checked the cables? Is there the possibility that there is a nasty noise source causing problems on the network? Or maybe a NIC is failing?

Installed TBB Meter which shows peaks of 76 Mbps when copying a 541MB file - here's the graph snapshot - the relevant one is the rightmost graph with lots of peaks and troughs.
Standard User MHC
(legend) Tue 31-May-11 00:41:08
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
That ties in with the speeds you were recording - 76 Mbps could be just the way it is reported and graphed, smoothing it out would show an average of around 40

Your big problem is going to be finding out what is causing the slowdown: Cable, NIC, interference, hard drive, CPU, router/switch





~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


M H C


taurus excreta cerebrum vincit
Standard User joconnell
(experienced) Tue 31-May-11 09:04:29
Print Post

Re: Measuring home network speed


[re: MHC] [link to this post]
 
In reply to a post by MHC:
Your big problem is going to be finding out what is causing the slowdown: Cable, NIC, interference, hard drive, CPU, router/switch

Yes, unfortunately frown

I can't see or think of any source of electrical interference, the NIC is a Realtek gigabit one integrated on the Asus P6TSE mobo of the newer PC (18 months old), the other NIC is a D-Link gigabit NIC on an old Dell PC circa 2001, so that could be an issue. I've read here and there that the mobo NICs are pretty poor though - would it be worth getting a PCI-e NIC?

Next things to try: use a fairly new laptop as a test endpoint to see if I get performance increase - thing is, the lappie only has fast ethernet but I guess if I see a performance increase over using the Dell, then that'd point to the older PC or the Dlink NIC being the issue.

Similarly, I could try using the 100Mbps switch built-in to my ADSL router (instead of the Netgear gigabit switch) and if performance increases there then that'd point to the switch being the weak link.

If I install a PCI NIC into the Asus mobo, I'm guessing that and the mobo NIC would operate on different IP addresses? If so I could do the test between those on the one PC, which would eliminate the other (older/slower) PCs as possible weak links...

Thanks for your feedback so far smile
Standard User MHC
(legend) Tue 31-May-11 09:48:03
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
From what you said: Your network is something like this.

My initial feeling is that the Netgear is not switching the traffic but sending it all to the ADSL router across the 100Mbit link where it is being switched.


Try it with both PCs connected to the router.





~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


M H C


taurus excreta cerebrum vincit
Standard User joconnell
(experienced) Tue 31-May-11 12:36:42
Print Post

Re: Measuring home network speed


[re: MHC] [link to this post]
 
In reply to a post by MHC:
From what you said: Your network is something like this.

My initial feeling is that the Netgear is not switching the traffic but sending it all to the ADSL router across the 100Mbit link where it is being switched.


Try it with both PCs connected to the router.

Good idea, I'll try that, I should get speeds closer to 100Mbps?.

Regarding the network diagram (I'm impressed), yes initially that is my setup, but I unplug the router from the Netgear gigabit switch once the DHCP server on the router has assigned the IP addresses to the PCs. The printer is connected to the router so it also isn't part of the switched network once I disconnect the router.

Maybe I should try assigning static IP addresses to the PCs and never have the netgear switch connected to the router before I test? Or perhaps the PCs would retain the previous IP addresses dished out by a previous DHCP session?
Standard User ggremlin
(member) Tue 31-May-11 13:06:32
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
Maybe I should try assigning static IP addresses to the PCs and never have the netgear switch connected to the router before I test?

or even use a crossover cable between the two pcs, with no switch
Standard User joconnell
(experienced) Tue 31-May-11 13:55:51
Print Post

Re: Measuring home network speed


[re: ggremlin] [link to this post]
 
In reply to a post by ggremlin:
Maybe I should try assigning static IP addresses to the PCs and never have the netgear switch connected to the router before I test?

or even use a crossover cable between the two pcs, with no switch

Good idea, but not sure how to configure static IP in WinXP - is it via the lmhosts file (to map host names to IP addresses) and the TCP/IP properties under the network connection dialog?
Anonymous
(Unregistered)Tue 31-May-11 15:35:38
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
In reply to a post by joconnell:
Good idea, but not sure how to configure static IP in WinXP - is it via the lmhosts file (to map host names to IP addresses) and the TCP/IP properties under the network connection dialog?

It's easy to configure static IP addresses under WinXP -It's done via the 'Internet Protocol (TCP/IP)' properties as you have already suggested.

N.B. You do not need a crossover Ethernet cable to connect 2 x 1Gb/s NICs directly to each other - ALL 1Gb/s NICs are Auto MDI/MDI-X capable.
Standard User joconnell
(experienced) Tue 31-May-11 17:07:24
Print Post

Re: Measuring home network speed


[re: Anonymous] [link to this post]
 
In reply to a post by Anonymous:
N.B. You do not need a crossover Ethernet cable to connect 2 x 1Gb/s NICs directly to each other - ALL 1Gb/s NICs are Auto MDI/MDI-X capable.

Thanks for the info, that'll save me time!
Standard User ggremlin
(member) Tue 31-May-11 17:09:40
Print Post

Re: Measuring home network speed


[re: Anonymous] [link to this post]
 
you may even be able to leave xp to get a 169.254.x.y address on each machine, and its autodiscovery to resolve names.
Standard User joconnell
(experienced) Tue 31-May-11 22:54:00
Print Post

Re: Measuring home network speed


[re: MHC] [link to this post]
 
In reply to a post by MHC:
From what you said: Your network is something like this.

My initial feeling is that the Netgear is not switching the traffic but sending it all to the ADSL router across the 100Mbit link where it is being switched.


Try it with both PCs connected to the router.

Did that and got 88Mbps consistently so it looks like the gigabit switch might be the problem. Next step, try direct connection between PCs.
Standard User MHC
(legend) Tue 31-May-11 23:46:07
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
That is about what I would expect.

No being familiar with Netgear but is there a possibility you have put it in "Bridge Mode" which means it just acts as a bridge between the Router and remote devices and does not switch locally?

Probably teaching you to suck eggs - but is DHCP off in the Netgear? Are all ports set for Gbit, Full Duplex?

Certainly if PC to PC gives say 250Mbit or above - there could be limitations on disk access which would slow the connection, then you can be confident that the NICs are in reasonable condition.

What if you power off everything, power up the Netgear, wait a minute or two, plug in te two PCs and boot them. What IP addresses do they get? Can they communicate and at what speed.





~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


M H C


taurus excreta cerebrum vincit
Anonymous
(Unregistered)Wed 01-Jun-11 01:14:48
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
Hmmm.

I have a GS605 too.

Or it would be more accurate to say "I had".

It replaced an FS108 8port 100Mbit switch and I had hoped it would increase the speed of large file copies on the LAN. Unfortunately I couldn't see any significant increase in speed, and did see a few unexplained oddities. I took the GS605 out and put the FS108 back, pending further investigation (which isn't likely to happen for a while).

Coincidence?

In an ideal world, two GBit-capable NICs and a suitable GBit switch will use "jumbo frames", which allows more data per ethernet packet and thus cuts down the proportion of overhead processing. So there should be a benefit from faster wires, and a benefit from fewer packets. Ideally. But if I remember rightly, for this to work requires MTU tweakery of the kind not much seen since BTwholesale built their core network wrong...
Anonymous
(Unregistered)Wed 01-Jun-11 01:18:51
Print Post

Re: Measuring home network speed


[re: MHC] [link to this post]
 
If it's the same as the GS605 I have, it's a simple unmanaged switch. There is nothing to configure in the switch, there is no DHCP server in the switch (or is that not what you were asking?).
Standard User joconnell
(experienced) Wed 01-Jun-11 10:25:13
Print Post

Re: Measuring home network speed


[re: ggremlin] [link to this post]
 
In reply to a post by ggremlin:
you may even be able to leave xp to get a 169.254.x.y address on each machine, and its autodiscovery to resolve names.
That did work but the PCs couldn't talk to each other, even with the firewalls off, so I ended up assigning static IPs.
Standard User joconnell
(experienced) Wed 01-Jun-11 10:35:20
Print Post

Re: Measuring home network speed


[re: MHC] [link to this post]
 
In reply to a post by MHC:
That is about what I would expect.

No being familiar with Netgear but is there a possibility you have put it in "Bridge Mode" which means it just acts as a bridge between the Router and remote devices and does not switch locally?

Probably teaching you to suck eggs - but is DHCP off in the Netgear? Are all ports set for Gbit, Full Duplex?

Certainly if PC to PC gives say 250Mbit or above - there could be limitations on disk access which would slow the connection, then you can be confident that the NICs are in reasonable condition.

What if you power off everything, power up the Netgear, wait a minute or two, plug in te two PCs and boot them. What IP addresses do they get? Can they communicate and at what speed.

The switch is a very simple 5 port home gigabit switch so there's no DHCP or bridge mode or any other sort of config, just pure "plug and play". With a direct connection from PC NIC to PC NIC, I copied a 1,164,853,580 byte file in 65 secs which I think equates to around 140 Mbps so the older computer's hard drive could be an issue.
Standard User joconnell
(experienced) Wed 01-Jun-11 10:50:51
Print Post

Re: Measuring home network speed


[re: Anonymous] [link to this post]
 
In reply to a post by Anonymous:
Hmmm.

I have a GS605 too.

Or it would be more accurate to say "I had".

It replaced an FS108 8port 100Mbit switch and I had hoped it would increase the speed of large file copies on the LAN. Unfortunately I couldn't see any significant increase in speed, and did see a few unexplained oddities. I took the GS605 out and put the FS108 back, pending further investigation (which isn't likely to happen for a while).

Coincidence?

Given it's price, I'm now thinking that the GS605 isn't going to set the world alight with it's performance which is probably fine in a situtation where high speed is less important and I can get away with speeds of around 150Mbps for streaming media or sharing an internet connection, but if I get a NAS drive, I'm going to need the benefits of gigabit ethernet which means speeds in the order of 700-800Mbps as reported by other posters in this thread.

As has been said, either the switch is the weak link, or the NICs or the PC to which I'm copying the test files. Given that I've taken the switch out of the equation with the direct PC-PC lan link, I'm starting to think that older test PC is just too slow and that's affecting the results - the dedicated LAN Speed Test tool does create a file on the endpoint PC so there's no getting away from hard drive performance.

What I will do is run a test with the newer PC as the endpoint and check the results there.
Standard User prlzx
(committed) Wed 01-Jun-11 12:10:05
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
Yes tests which involve hard disk access that can be a limiting factor, at least until SSDs are the norm, though stuff like iperf should indeed work, not had cause to try it myself so don't know about its options.

With newer NICs they can offload some of the work from the CPU. With older NICs a slower CPU could be tied up servicing the network card (this is where jumbo frames can help, up to a factor of 6 for the larger packets).

To do timed tests with files you could create RAMdisks on both machines (certainly with unix-like boxes some tmpfs sized 0.5 to 1GB would do) - duplicating various sizes of TBB test files makes a suitable collection for this purpose smile

Whereas to see if the packet rate is affecting the CPU on the older machine you could watch the appropriate tab in task manager / process explorer / system monitor / other utility of your choice throughout the tests.



prompt $P - Invalid drive specification - Abort, Retry, Fail? $G
prlzx on n e w n e t Max ADSL
Standard User gmoorc
(member) Wed 01-Jun-11 12:32:34
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
In reply to a post by joconnell:
Given it's price, I'm now thinking that the GS605 isn't going to set the world alight with it's performance


I had nothing but trouble with my GS605.
Would regularly drop back to fast ethernet on my PC connection although the link to my NAS was OK at Gigabit.
When it was connected as Gigabit I never got more than around 300Mbps speeds.

I have now switched to a Cisco home switch which is rock solid and runs around 700Mbps.
Standard User prlzx
(committed) Wed 01-Jun-11 12:52:05
Print Post

Re: Measuring home network speed


[re: Anonymous] [link to this post]
 
Jumbo frames generally depends on the drivers for the network card - if the adapter properties has some kind of advanced tab or a separate utility (e.g. the Intel Pro stuff).

With a modern operating system there is a default MTU for the interface (usually 1500, but it can take into account if DHCP gives out a different value).
However MTUs to other destinations are reduced on the fly (by PMTUD) as long as the intervening routes allow ICMP (pings and such).

What this means is if all devices on the LAN can cope with jumbo frames you can try it and a modern OS should end up using the larger MTU on the LAN but a smaller value when talking to the internet (*).

You'll want the router to officially support jumbo frames on the LAN interface as it should by design fragment them for a smaller MTU on the WAN interface, even if a device mistakenly tries to send oversize packets towards the internet.

---

There are at least 3 factors that might hamper automatic MTU detection / NIC performance:

- If the MTU has been manually tweaked in the past (if there was only a single compromise MTU) the automatic behaviour might have been turned off.

- If a destination site's network admins have block pings (misguided ideas about security) it might take longer to find a working MTU for that path.

- If you have ever used the Windows "detect and repair" network troubleshooter that turns off certain optimisations (favours conservative over speed).



prompt $P - Invalid drive specification - Abort, Retry, Fail? $G
prlzx on n e w n e t Max ADSL
Standard User joconnell
(experienced) Wed 01-Jun-11 14:34:06
Print Post

Re: Measuring home network speed


[re: prlzx] [link to this post]
 
You'll want the router to officially support jumbo frames on the LAN interface as it should by design fragment them for a smaller MTU on the WAN interface, even if a device mistakenly tries to send oversize packets towards the internet

Well the router is a Thomson TG585 v7 with 100Mbps ethernet ports, so jumbo frames don't apply do they? The Netgear GS605 switch (into which the router is plugged in order to share the internet connection) has no config options and the gigabit NICs have jumbo frames disabled (because there are also 100Mbps devices on the network)
Standard User Als
(experienced) Wed 01-Jun-11 14:52:22
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
Did that and got 88Mbps consistently so it looks like the gigabit switch might be the problem. Next step, try direct connection between PCs.


That sounds like the same problem I had with a GS608.
I replaced it with a cheap Zyxel 8 port switch & Win 7 consistently reports good speeds around 90 - 100MB/sec.

Als

BE Value
Standard User joconnell
(experienced) Wed 01-Jun-11 15:19:33
Print Post

Re: Measuring home network speed


[re: Als] [link to this post]
 
In reply to a post by Als:
That sounds like the same problem I had with a GS608.
I replaced it with a cheap Zyxel 8 port switch & Win 7 consistently reports good speeds around 90 - 100MB/sec.

I take it you mean 90-100 megabytes per sec...
Standard User prlzx
(committed) Wed 01-Jun-11 16:18:05
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
In reply to a post by joconnell:
router to ... support jumbo frames on the LAN interface

Well the router is a Thomson TG585 v7 with 100Mbps ethernet ports, so jumbo frames don't apply do they? The Netgear GS605 switch (into which the router is plugged in order to share the internet connection) has no config options and the gigabit NICs have jumbo frames disabled (because there are also 100Mbps devices on the network)

Yes for your setup you would probably only use jumbo frames temporarily as part of testing transfers. There are consumer routers with gigabit interfaces so I left that in for the more general case.

I had looked into jumbo frames for a work situ with a similar result that it's not yet convenient to run the jumbo and non-jumbo (or 100M) stuff on separate networks (or VLANs). The non-jumbo stuff included things like shared printers and even some Mac laptops which I found a surprising omission.

I'm guessing a mixed setup might work but with more retransmissions so not necessarily helping.



prompt $P - Invalid drive specification - Abort, Retry, Fail? $G
prlzx on n e w n e t Max ADSL
Standard User joconnell
(experienced) Thu 02-Jun-11 00:40:37
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
In reply to a post by joconnell:
Next step, try direct connection between PCs.

Okay, I've connected both PCs directly with a 2m Cat5e cable, assigned static IP addresses and run speed tests using LAN Speed Test, and I get speeds of around 140Mbps.
Using Ixia QTest reports 200Mbps. With 8KB jumbo frames enabled on both NICs, that result is doubled, but I've left jumbo frames off as the network is a mix of gigabit and 100Mbps ethernet.
Running iPerf reports speeds of 9.5Mbps!
A 2 gigabyte file copy from the slower PC to the faster PC takes 1 min 45 secs - I've run the test in that direction as I've reasoned that reading a file from a slow machine and writing to a fast machine reduces skewed results resulting from a slower system.

Now given that the network consists of one 2m patch cable (not a crossover cable) linking 2 gigabit NICs, I'm at a loss to explain why performance is so poor. Others have reported speeds of 700Mbps or more on gigabit networks. How are others testing their networks?

The cable is brand new and I've tried a few cables, no difference to the test results. Unless one or both of the NICs cannot perform to gigabit speeds due to a fault or just shortcomings of the NIC?

Has anyone any more suggestions?

Edited by joconnell (Thu 02-Jun-11 00:46:10)

Standard User Als
(experienced) Thu 02-Jun-11 08:01:33
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
I take it you mean 90-100 megabytes per sec...


I've just copied two video files acroos my lan. The first file was 3.83GB & took 53 seconds. The second was 3.95Gb & took 47 secs.

Snapshot 1
Snapshot 2

Both pc's have Gigabyte motherboards with onboard gigabit nics & both are running windows 7.

Als

BE Value

Edited by Als (Thu 02-Jun-11 08:19:33)

Standard User Als
(experienced) Thu 02-Jun-11 09:05:58
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
Running iPerf reports speeds of 9.5Mbps!
A 2 gigabyte file copy from the slower PC to the faster PC takes 1 min 45 secs - I've run the test in that direction as I've reasoned that reading a file from a slow machine and writing to a fast machine reduces skewed results resulting from a slower system.


Thats similar to speeds I was getting with XP & ide drives. I couldn't pinpoint the problem but when I installed Win 7, speeds immediately improved to 50Mbs. This is on an old Abit socket 754 with onboard gb nic.

Quad core Gigabyte to Abit socket 754

Als

BE Value
Standard User joconnell
(experienced) Thu 02-Jun-11 11:31:57
Print Post

Re: Measuring home network speed


[re: Als] [link to this post]
 
In reply to a post by Als:
That's similar to speeds I was getting with XP & ide drives. I couldn't pinpoint the problem but when I installed Win 7, speeds immediately improved to 50Mbs. This is on an old Abit socket 754 with onboard gb nic.

Quad core Gigabyte to Abit socket 754

Do you have jumbo frames enabled for your gigabit nics and is every device on your LAN gigabit?

So was Win XP the weak link? Do you have a win7 PC with IDE drives which now performs much better? What's the spec of the machine you were copying from/to?

Edited by joconnell (Thu 02-Jun-11 12:15:20)

Anonymous
(Unregistered)Thu 02-Jun-11 12:47:26
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
Odd, isn't it. Tests measure what tests measure, and may or may not reflect the ultimate capability of two NICs in two computers.

If you can find some relevant Linux tests you might like to cut out the Windows IP stack and see if that sheds any more light.

"A 2 gigabyte file copy from the slower PC to the faster PC takes 1 min 45 secs - I've run the test in that direction as I've reasoned that reading a file from a slow machine and writing to a fast machine reduces skewed results resulting from a slower system."

Rather than assuming, you might want to actually try it, if it's not going to be difficult.

If the slow computer is doing the reading, surely that WILL limit the transfer speed as there is nothing to send if there's no data available? If the slow computer is doing the writing, it's not so much a bottleneck as data it receives can in principle be cached in memory before being written to disk? Not saying that's 100% what happens, but I believe NTFS on Windows does by default write-behind like that, because it does provide some apparent performance gain when writing. The downside of write-behind like that is that data in memory but not on disk is vulnerable to being lost if Windows crashes or if power fails, etc.
Standard User Als
(experienced) Thu 02-Jun-11 13:00:22
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
No, I don't have jumbo frames enabled as we have a couple of laptops with 100Mb cards.

I'm quite sure XP was part of the problem. I could get good rates between two Win7 machines but not from Win7 to XP. The Win7 machines have sata drives whereas the XP machine was all IDE. As soon as I installed Win7 on the old XP box the nw speed improved by at least 5x.

The old XP, now Win7, box is Abit NF8, 3000 Athlon, onboard Realtek gb nic,1G ram, 3 IDE hard drives, 1 IDE dvd.
The other Win7 boxes are Gigabyte, onboard Realtek gb nics, Intel cpus 4 & 6GB rams, all sata drives.

When I had XP on the Gigabyte boxes I got reasonably good lan speeds but my GS608 was playing around on a couple of ports. Replacing it with the Zyxel switch cured the slow speeds. But I still had the slow speeds to the Abit NF8 Athlon box which was also on XP. After trying everything I could think of I decided to try Win7. There was an immediate improvement. Its not perfect & I'm guessing that its either the pci bus or ide drives that are the limiting factor.

My daughter's bf checked my ethernet cables with very expensive pro test gear & found nothing wrong.

I've got another old XP box I could fire up & do some tests. Its a cheap & nasty Elite or ECS board, iirc. I have a netgear 311 gb nic that I can try oout & see what speeds I get.

Als

BE Value
Standard User joconnell
(experienced) Thu 02-Jun-11 13:31:14
Print Post

Re: Measuring home network speed


[re: Als] [link to this post]
 
In reply to a post by Als:
No, I don't have jumbo frames enabled as we have a couple of laptops with 100Mb cards.

I'm quite sure XP was part of the problem. I could get good rates between two Win7 machines but not from Win7 to XP. The Win7 machines have sata drives whereas the XP machine was all IDE. As soon as I installed Win7 on the old XP box the nw speed improved by at least 5x.

The old XP, now Win7, box is Abit NF8, 3000 Athlon, onboard Realtek gb nic,1G ram, 3 IDE hard drives, 1 IDE dvd.
The other Win7 boxes are Gigabyte, onboard Realtek gb nics, Intel cpus 4 & 6GB rams, all sata drives.

When I had XP on the Gigabyte boxes I got reasonably good lan speeds but my GS608 was playing around on a couple of ports. Replacing it with the Zyxel switch cured the slow speeds. But I still had the slow speeds to the Abit NF8 Athlon box which was also on XP. After trying everything I could think of I decided to try Win7. There was an immediate improvement. Its not perfect & I'm guessing that its either the pci bus or ide drives that are the limiting factor.

My daughter's bf checked my ethernet cables with very expensive pro test gear & found nothing wrong.

I've got another old XP box I could fire up & do some tests. Its a cheap & nasty Elite or ECS board, iirc. I have a netgear 311 gb nic that I can try oout & see what speeds I get.

Thanks for all your info.

The only things I can think is that I should use a crossover cable rather than a patch cable to connect the two PCs, though with gigabit NICs I've been told a crossover cable isn't needed. I think I saw a van from a datacomms firm parked at a house near me, I might have a word and see if I can get my home network checked out, at least that way I'll know it's alright, which is the point of the testing I'm doing now.

But ultimately, the best bet is to test between a couple of modern Win 7 PCs, or as someone suggsted, use Linux to test.
Standard User joconnell
(experienced) Thu 02-Jun-11 13:35:09
Print Post

Re: Measuring home network speed


[re: Anonymous] [link to this post]
 
In reply to a post by Anonymous:
Odd, isn't it. Tests measure what tests measure, and may or may not reflect the ultimate capability of two NICs in two computers.

If you can find some relevant Linux tests you might like to cut out the Windows IP stack and see if that sheds any more light.

That's a good idea, I wonder if I could run tests off a CD booted version of Ubuntu or whatever? Can Linux access the local windows disks?
Standard User joconnell
(experienced) Thu 02-Jun-11 20:29:27
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
I don't think I'm going to get much further with my testing until I can get another up to date Win7 PC with a gigabit NIC. Here's my current results of testing by copying two files from one 10 year old PC to an 18 month old PC, both with gigabit NICs.

Text
1
23
45
67
89
1011
direct connect with patch cable
2,090,110,986 in 1.35 mins [168 Mbps]1,482,419,895 in 1.09 mins [164 Mbps]
------------------------------------------------------------------------------------------connected via switch with router (100Mbps) plugged into switch
2,090,110,986 in 1.40 mins [160 Mbps]1,482,419,895 in 1.12 mins [158 Mbps]
------------------------------------------------------------------------------------------connected via switch with no router plugged in
2,090,110,986 in 1.40 mins [160 Mbps] 1,482,419,895 in 1.09 mins [164 Mbps]


At this point the best I can do is to establish whether the tested speeds on my home gigabit LAN are close to the above results, that way I'll know that at least the LAN cabling is terminated okay at the RJ45 modules and at the patch panel.

Thanks for all your help so far.
Anonymous
(Unregistered)Fri 03-Jun-11 09:01:38
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
"Can Linux access the local windows disks? "

Any sensible recent Linux can access FAT, FAT32, and NTFS, for both read and write.

Ideally you'd find a reliable pair of test programs that don't involve the disk, if what you want to measure is NIC/network performance.

Perhaps a Linux expert could suggest a LiveCD to choose, and/or suggest whether simple test programs (not necessarily ftp, samba, etc) are readily available.

But if Windows is your chosen environment, testing under Linux may be a distraction, as it probably won't help improve Windows performance (unless the Linux tests help discover and fix a problem with the NICs or the network).

Best of luck anyway.
Standard User joconnell
(experienced) Fri 03-Jun-11 22:07:28
Print Post

Re: Measuring home network speed


[re: joconnell] [link to this post]
 
After some further investigation I believe the issue to be with the old Dell PC - when copying files from it or to it, the CPU usage maxes out at 100% so it looks like it can't cope with the overhead. NIC usage peaks at 26% but is 17% average. On the other hand, the newer PC CPU usage maxes out at 25%.

So I've given up speed testing for now, at least until I get another newer gigabit NIC-equipped PC.

The good news (I think) is that copying from one PC to the other takes the same time across my home LAN as it does with the PCs directly connected without a switch or anything else on the network, so hopefully when I do test speeds with another faster and more modern PC, I'll be getting the sort of speeds reported by others in this thread, in the order of 700-900 Mbps.

Thanks for all your help and advice, I've learned quite a bit from your feedback.
Pages in this thread: 1 | 2 | 3 | 4 | 5 | >> (show all)   Print Thread

Jump to