Hi all! Got the following problem. Have a home network, with 3 PC's running SuSE (2 times 9.3 and 1 9.1) and a hardware router linking it all together. One of the PC's is to be a server, serving files mainly. Everything installs just fine, I get access to the server and all (both with NFS and Samba). Downloads from the server down to the clients is as fast as it gets on 100Mb ethernet, but upload TO the server is extremely slow and results in a lot of errors on the server. See below: gothix:~ # ifconfig eth0 Link encap:Ethernet HWaddr 00:40:CA:17:7F:0B inet addr:192.168.0.69 Bcast:192.168.0.255 Mask:255.255.255.0 UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1 RX packets:82728 errors:22059 dropped:0 overruns:0 frame:33165 TX packets:75140 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:98389465 (93.8 Mb) TX bytes:8641743 (8.2 Mb) Interrupt:10 Base address:0x1080 lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:74 errors:0 dropped:0 overruns:0 frame:0 TX packets:74 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:5184 (5.0 Kb) TX bytes:5184 (5.0 Kb) I have tried everything that google can throw at me, but no results. Any idea what might be the problem here? If I convert one of the other PC's into a server and the "server" into a client, same result: from the temporary server to the temporary client (the real server): dead slow. The server is fast otherwise, can download from the internet and all without a hitch. Any help highly appreciated. Oh, the nic is a onboard 3Com gothix:~ # ifstatus eth0 eth0 device: 3Com Corporation 3c905B 100BaseTX [Cyclone] (rev 30) eth0 DHCP client (dhcpcd) is running IPADDR=192.168.0.69 NETMASK=255.255.255.0 GATEWAY=192.168.0.1 DNS=192.168.0.1,192.168.0.1 Cheers, Frits
It appears you are saying that your "server" machine has trouble receiving (no matter if it is acting as a server or a client), if I understand your statement of the problem. At this point, I would suspect hardware. Some things to try... 1) have you tried a different _client_ machine to send to the server? Are the results the same? If not, problem is on "server" machine's RCV port. 2) try switching the cable around (plug the ends of the cables into the opposite machine). This could point to the cable if your "server" machine suddenly starts RCV'ing ok, but the client now gets RCV errors. -or- (better) 3) Try a different "known good cable". 4) Do you have any alternate ethernet cards you can try? -- Perhaps in the "Server" machine you have 2 network cards? Can you temporarily reconfiguring to do the tests with the other network card, or do you have a spare around to try to substitute in? 5) If both machines are linux, you can rule out file-system interactions by using the "nttcp" package. Run on 1 machine (artibrary client) with "nttcp -i" and then on other machine -- output looks roughly like: machine1:/> nttcp -T machine2 Bytes Real s CPU s Real-MBit/s CPU-MBit/s Calls Real-C/s CPU-C/s l 8388608 0.71 0.04 94.7974 1560.9254 2048 2892.99 47635.7 1 8388608 0.71 0.12 94.0853 568.8060 5809 8144.10 49236.3 6) if one is a windows machine, you might use the "cygwin" package to load "rsh" or "ssh". You could xfer files with rsync (which can tell you speed), or use "dd" into a pipe. Be aware that rsh is insecure and should be made visible to "the outside world", and 2) ssh has performance overhead due to it's encryption. 7) Make sure network card in suspect machine is "reseated" -- possibly even try a different slot -- not likely to change things, but I've seen it happen. Good luck... Linda Walsh Frits Spieker wrote:
gothix:~ # ifconfig eth0 Link encap:Ethernet HWaddr 00:40:CA:17:7F:0B inet addr:192.168.0.69 Bcast:192.168.0.255 Mask:255.255.255.0 UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1 RX packets:82728 errors:22059 dropped:0 overruns:0 frame:33165 TX packets:75140 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:98389465 (93.8 Mb) TX bytes:8641743 (8.2 Mb) Interrupt:10 Base address:0x1080
lo Link encap:Local Loopback inet addr:127.0.0.1 Mask:255.0.0.0 UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:74 errors:0 dropped:0 overruns:0 frame:0 TX packets:74 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:5184 (5.0 Kb) TX bytes:5184 (5.0 Kb)
I have tried everything that google can throw at me, but no results.
Any idea what might be the problem here?
If I convert one of the other PC's into a server and the "server" into a client, same result: from the temporary server to the temporary client (the real server): dead slow.
On Sunday, 12. June 2005 21:36, Linda Walsh wrote:
It appears you are saying that your "server" machine has trouble receiving (no matter if it is acting as a server or a client), if I understand your statement of the problem.
You understand correctly ;-)
At this point, I would suspect hardware. Some things to try... 1) have you tried a different _client_ machine to send to the server? Are the results the same? If not, problem is on "server" machine's RCV port.
I tried all the easy things first (other cable, setting MTU etc) and then went for the hardware. Indeed it seems that the onboard NIC was at fault. Plugged in a NIC that I still had lying around, reconfigured and presto, We're in business. Thanks! (you too Ken) Cheers, Frits
On Sun, 2005-06-12 at 18:18 +0200, Frits Spieker wrote:
Hi all!
Got the following problem.
Have a home network, with 3 PC's running SuSE (2 times 9.3 and 1 9.1) and a hardware router linking it all together.
One of the PC's is to be a server, serving files mainly.
Everything installs just fine, I get access to the server and all (both with NFS and Samba).
Downloads from the server down to the clients is as fast as it gets on 100Mb ethernet, but upload TO the server is extremely slow and results in a lot of errors on the server.
See below: gothix:~ # ifconfig eth0 Link encap:Ethernet HWaddr 00:40:CA:17:7F:0B inet addr:192.168.0.69 Bcast:192.168.0.255 Mask:255.255.255.0 UP BROADCAST NOTRAILERS RUNNING MULTICAST MTU:1500 Metric:1 RX packets:82728 errors:22059 dropped:0 overruns:0 frame:33165
The RX is showing 22059 errors. You might check the MTU size and if it is 1500 change it to 1492. this may clear up the TX errors.
TX packets:75140 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:98389465 (93.8 Mb) TX bytes:8641743 (8.2 Mb) Interrupt:10 Base address:0x1080
I have tried everything that google can throw at me, but no results.
Any idea what might be the problem here?
If I convert one of the other PC's into a server and the "server" into a client, same result: from the temporary server to the temporary client (the real server): dead slow.
This confirms that there is a problem with this machine, the original server. -- Ken Schneider UNIX since 1989, linux since 1994, SuSE since 1998 "The day Microsoft makes something that doesn't suck is probably the day they start making vacuum cleaners." -Ernst Jan Plugge
participants (3)
-
Frits Spieker -
Ken Schneider -
Linda Walsh