100Mb lan tops out at 1.2Mb ?

If two machines running 100 Megabit nics are plugged
into a 10/100 Megabit switch, does this mean the transfer
speeds should be at or near 100 Megabit/second in
optimum conditions?

What would be some things to check that would
increase that 6MB/s speed?

Thanks,
chubbs


------------ ftp to server and retrieve file
----------------------------
ftp> get a_army.zip
local: a_army.zip remote: a_army.zip
227 Entering Passive Mode (192,168,1,1,128,17)
150 Opening BINARY mode data connection for 'a_army.zip' (769199104
bytes).
18% |*******************
| 134 MB 6.09 MB/s
01:38 ETA^


------------- server nic is running 100Mb
-------------------------------
[root@kong: ~ 1]$ mii-tool
eth0: negotiated 100baseTx-FD, link ok



--------------- client nic is at 100Mb
-------------------------------------
[root@fungus /tmp]# ethtool eth0
Settings for eth0:
Supported ports: [ TP MII ]
Supported link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
Supports auto-negotiation: Yes
Advertised link modes: 10baseT/Half 10baseT/Full
100baseT/Half 100baseT/Full
Advertised auto-negotiation: Yes
Speed: 100Mb/s
Duplex: Full
Port: MII
PHYAD: 32
Transceiver: internal
Auto-negotiation: on
Supports Wake-on: pumbg
Wake-on: d
Current message level: 0x00000007 (7)
Link detected: yes

On 10/07/05 14:22, (E-Mail Removed) wrote:

> ------------ ftp to server and retrieve file
> ----------------------------
> ftp> get a_army.zip
> local: a_army.zip remote: a_army.zip
> 227 Entering Passive Mode (192,168,1,1,128,17)
> 150 Opening BINARY mode data connection for 'a_army.zip' (769199104
> bytes).
> 18% |*******************
> | 134 MB 6.09 MB/s
> 01:38 ETA^
>

Get faster disks or faster processors. Anyway you are over 50% of the
band and it is not too bad.

Ciao
Giovanni
--
A computer is like an air conditioner,
it stops working when you open Windows.
Registered Linux user #337974 <http://counter.li.org/>

<(E-Mail Removed)> wrote in message
news:(E-Mail Removed) oups.com...

> If two machines running 100 Megabit nics are plugged
> into a 10/100 Megabit switch, does this mean the transfer
> speeds should be at or near 100 Megabit/second in
> optimum conditions?

Yes.

> What would be some things to check that would
> increase that 6MB/s speed?

You understand that MB/s is *NOT* megabits per second, right? You should
also know that you should not expect effective data speeds to equal line
speeds.

DS

(E-Mail Removed) wrote:

> If two machines running 100 Megabit nics are plugged
> into a 10/100 Megabit switch, does this mean the transfer
> speeds should be at or near 100 Megabit/second in
> optimum conditions?
>
> What would be some things to check that would
> increase that 6MB/s speed?

6 Mega Bytes per second (48 Megabits) + overhead is not unreasonable for a
100 Mbit network, although I'd expect a little more on an unloaded network.
If one of the machines is slower than a Pentium 200 or has a slow disk, it
won't be able to use the full bandwidth available.

(E-Mail Removed) writes:

>If two machines running 100 Megabit nics are plugged
>into a 10/100 Megabit switch, does this mean the transfer
>speeds should be at or near 100 Megabit/second in
>optimum conditions?

100Mbit= 12MB.
So 6MB/s is half the theoretical max rate. If the switch is not great it
could limit the transfer. If the computer has to do disk io it could limit
the rate at which it could feed the pipe. Ie, IF you get 6MB/s I would not
worry too much.

>What would be some things to check that would
>increase that 6MB/s speed?

(E-Mail Removed) wrote:
> What would be some things to check that would increase that 6MB/s
> speed?

First, test with something that does not involve disc I/O - www.netperf.org

Second, make sure you are not maxing-out _any_ of the CPU(s) on either
side.

Third, make sure you have a "large enough" TCP window.

Fourth, make sure there are no TCP retransmissions

Fifth, make sure there isn't a duplex mismatch between your hosts
and/or the switch.

rick jones

How Autoneg is supposed to work:

When both sides of the link are set to autoneg, they will "negotiate"
the duplex setting and select full duplex if both sides can do
full-duplex.

If one side is hardcoded and not using autoneg, the autoneg process
will "fail" and the side trying to autoneg is required by spec to use
half-duplex mode.

If one side is using half-duplex, and the other is using full-duplex,
sorrow and woe is the usual result.

So, the following table shows what will happen given various settings
on each side:

Auto Half Full

Auto Happiness Lucky Sorrow

Half Lucky Happiness Sorrow

Full Sorrow Sorrow Happiness

Happiness means that there is a good shot of everything going well.
Lucky means that things will likely go well, but not because you did
anything correctly Sorrow means that there _will_ be a duplex
mis-match.

When there is a duplex mismatch, on the side running half-duplex you
will see various errors and probably a number of late collisions. On
the side running full-duplex you will see things like FCS errors.
Note that those errors are not necessarily conclusive, they are simply
indicators.

--
oxymoron n, commuter in a gas-guzzling luxury SUV with an American flag
these opinions are mine, all mine; HP might not want them anyway...
feel free to post, OR email to rick.jones2 in hp.com but NOT BOTH...

Yes, but that throughput inlcudes TPC/IP overhead, packet information
etc. When copying by FTP, you get a report of the number of bytes of
the file transferred, NOT total data throughput. However, even so, you
can still get a little over 10,000 Kb/s


I find the quality of the network cable and switch can have an impact.
On my simple network, linking two PC's via a switch gives much better
transfer rates then directly.
You may want to try hdparm and check how to optimize your hard drive.
Quite possibly your system is too busy writing/reading to transfer at
full speed. Enabling a few of these flags, if not already enabled may
make a noticable difference.



Unruh wrote:
> (E-Mail Removed) writes:
>
>
>>If two machines running 100 Megabit nics are plugged
>>into a 10/100 Megabit switch, does this mean the transfer
>>speeds should be at or near 100 Megabit/second in
>>optimum conditions?
>
>
> 100Mbit= 12MB.
> So 6MB/s is half the theoretical max rate. If the switch is not great it
> could limit the transfer. If the computer has to do disk io it could limit
> the rate at which it could feed the pipe. Ie, IF you get 6MB/s I would not
> worry too much.
>
>
>>What would be some things to check that would
>>increase that 6MB/s speed?
>
>
>
>

Snowbat wrote:
<snip>

> If one of the machines is slower than a Pentium 200 or has a slow disk, it
> won't be able to use the full bandwidth available.

Ahh.. I bet you nailed it then. The host is a 266 PII.

Thanks

David Schwartz wrote:
<snip>

> You understand that MB/s is *NOT* megabits per second, right? You should
> also know that you should not expect effective data speeds to equal line
> speeds.

Yup - big B little b. Is it unreasonable to assume a 100mb
nic/switch/etc will never get 100mb/s under reasonable conditions?

(E-Mail Removed) wrote:

> David Schwartz wrote:
> <snip>

>> You understand that MB/s is *NOT* megabits per second, right?
>> You should also know that you should not expect effective data
>> speeds to equal line speeds.

> Yup - big B little b. Is it unreasonable to assume a 100mb
> nic/switch/etc will never get 100mb/s under reasonable conditions?

Actually, it always gets 100mb/s when it's sending or receiving data.

Seriously, you should get pretty close unless you are limited by disk or
CPU. Note that this still won't mean that your data throughput will ever get
to 100mb/s, because there's headers, encoding overhead, acknowledgements,
and so on.

DS