Performance of 3C575TX query
Stephen J. Gowdy
gowdy@mh1.lbl.gov
Mon Oct 5 18:56:16 1998
Hi Donald,
I thought I'd try the .99G performance with the above card on a
100Mbps network. I used ftp between a large Digital UNIX server with RAID
striped disks and my Dell Latitude CPi 266XT. Connected to a 10Mbps hub
very close to the server I get the following (I just realised I should
have renamed this test file, please forget the name it isn't important);
ftp> get CXX610V27.tar.gz
local: CXX610V27.tar.gz remote: CXX610V27.tar.gz
200 PORT command successful.
150 Opening BINARY mode data connection for CXX610V27.tar.gz
(128.3.5.227,8778) (16901331 bytes).
226 Transfer complete.
16901331 bytes received in 18.3 secs (9e+02 Kbytes/sec)
ftp> put CXX610V27.tar.gz
local: CXX610V27.tar.gz remote: CXX610V27.tar.gz
200 PORT command successful.
150 Opening BINARY mode data connection for CXX610V27.tar.gz
(128.3.5.227,8788).
226 Transfer complete.
16901331 bytes sent in 17.4 secs (9.5e+02 Kbytes/sec)
However, when I use 100Mbps connection to the same hub as the
machine I get the following;
ftp> put CXX610V27.tar.gz test
local: CXX610V27.tar.gz remote: test
200 PORT command successful.
150 Opening BINARY mode data connection for test (128.3.5.227,8762).
226 Transfer complete.
16901331 bytes sent in 6.04 secs (2.7e+03 Kbytes/sec)
ftp> get CXX610V27.tar.gz test
local: test remote: CXX610V27.tar.gz
200 PORT command successful.
150 Opening BINARY mode data connection for CXX610V27.tar.gz
(128.3.5.227,8772) (16901331 bytes).
receive aborted
waiting for remote to finish abort
426 Transfer aborted. Data connection closed.
226 Abort successful
341640 bytes received in 22.5 secs (15 Kbytes/sec)
So I aborted the connection as it was taking so long. So to
summerise, with 10baseT its just under 1Mbyte per second and with
100baseTX it is 3Mbytes per second transmit and 15kbytes per second
recieve. netstat -i shows a lot of RX errors which accumulate only when
on the 100baseTX;
Kernel Interface table
Iface MTU Met RX-OK RX-ERR RX-DRP RX-OVR TX-OK TX-ERR TX-DRP TX-OVR
Flags
lo 3584 0 22680 0 0 0 22680 0 0 0
BLRU
eth0 1500 0 161125 1062 0 1062 88808 0 0 0
BNRU
I'm not sure if netstat -s tells you anything but here is the
output from that;
bash# netstat -s
Ip:
140709 total packets received
0 with invalid headers
0 with invalid addresses
0 forwarded
0 with unknown protocol
0 incoming packets discarded
23109 incoming packets delivered
126889 requests sent out
0 outgoing packets dropped
797 dropped because of missing route
0 fragments dropped after timeout
0 reassemblies required
0 packets reassembled ok
0 packet reassembles failed
0 fragments received ok
0 fragments failed
0 fragments created
Icmp:
557 ICMP messages received
0 input ICMP message failed.
ICMP input histogram:
destination unreachable: 140
echo replies: 417
120 ICMP messages sent
0 ICMP messages failed
ICMP output histogram:
destination unreachable: 120
Tcp:
1170 active opens
0 passive opens
16 failed connection attempts
4 connection resets received
3 connections established
117410 segments received
121781 segments send out
215 segments retransmited
Udp:
2337 packets received
475 packets to unknown port received.
0 packet receive errors
2277 packets send
So I suppose I have three questions;
1) Should I expect 1Mbyte/s for 10baseT and 10Mbytes/s for
100baseTX?
2) Is there a better way to measure actual performance?
3) Is this a known problem?
I don't really need the performance just now so don't feel under
any presure to resolve this for me, I'd just like to know about this
issue.
regards,
Stephen.
--
/------------------------------+-=-=-=-=-+-------------------------\
|Stephen J. Gowdy |A4000/040| Mail Stop 50A-2160, LBL, |
|http://www.ph.ed.ac.uk/~gowdy/ | 1GB HD| 1 Cyclotron Rd, Berkeley,|
| |20MB RAM| CA 94720, USA |
|InterNet: SGowdy@lbl.gov |3.4xCDROM| Tel: +1 510 495 2796 |
\------------------------------+-=-=-=-=-+-------------------------/