Re: [expert] Network problem
Richard Bown (QMW) wrote: Hi all, It would help Wolfgang if you could send the routing table on both machines ! From Wolfgang's *first* post: Desktop: 217.5.98.14 0.0.0.0 255.255.255.255 UH 0 0 0 ppp0 10.0.0.00.0.0.0 255.255.255.0 U 0 0 0 eth0 192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0 eth1 127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo 0.0.0.0 217.5.98.140.0.0.0 UG 0 0 0 ppp0 Notebook: 192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 127.0.0.0 0.0.0.0 255.0.0.0U 0 0 0 lo 0.0.0.0 192.168.0.10.0.0.0 UG 0 0 0 eth0 Pierre Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Richard Bown (QMW) wrote: [snip] I'm waiting to hear back from Wolfgang on the list of things to verify... Until a ping -br broadcast which bypasses the routing table actually makes it to the other machine (basic connectivity), routing tables changes are not going to help (unless the network code is buggy). Cheers, Pierre Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
RE: [expert] Network problem
Not so sure on that Pierre, it will get a reply from what it can see, bypassing the routing table will only get a ping back from the first device. if 100 machines are connected to the port and the port has the same address as the host then all hell breaks loose, but if the host address is different from the port address, only the port responds, because the host dos'nt know whats the otherside of the port unless its told.. ie. the routing table richard -Original Message- From: Pierre Fortin [mailto:[EMAIL PROTECTED]] Sent: Wednesday, November 28, 2001 5:22 PM To: [EMAIL PROTECTED] Subject: Re: [expert] Network problem Richard Bown (QMW) wrote: [snip] I'm waiting to hear back from Wolfgang on the list of things to verify... Until a ping -br broadcast which bypasses the routing table actually makes it to the other machine (basic connectivity), routing tables changes are not going to help (unless the network code is buggy). Cheers, Pierre Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Am Mittwoch, 28. November 2001 17:22 schrieben Sie: Richard Bown (QMW) wrote: [snip] I'm waiting to hear back from Wolfgang on the list of things to verify... Until a ping -br broadcast which bypasses the routing table actually makes it to the other machine (basic connectivity), routing tables changes are not going to help (unless the network code is buggy). That's the point. Unless I can get a connection from notebook to desktop I don't bother about routes. I ripped off the NIC which was configured as eth1 b/c it was a n old ISA card and I mistrust them. Also the MCC kept saying You only have one NIC although both showed up in the network section as up. So much to the gui tools. Right now I got a cheap NIC called Netgear FA311 Fast Ethernet PCI Card. Problem is, it comes with Linux drivers (!) but only for Red Hat 6.x and the given files will not compile b/c the makefile is counting on kgcc. Duh... So I cannot build the module. Harddrake doesn't know it, it says it's a NAtional Semiconductor Product and that it is unknown wobo (really getting in to it!) Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
OK, I'm answering to my own mail b/c there's an update: Right now I got a cheap NIC called Netgear FA311 Fast Ethernet PCI Card. Problem is, it comes with Linux drivers (!) but only for Red Hat 6.x and the given files will not compile b/c the makefile is counting on kgcc. Duh... So I cannot build the module. Harddrake doesn't know it, it says it's a NAtional Semiconductor Product and that it is unknown Through google I found the hint to use the natsemi module and it works. I have a working connection between notebook and desktop! Seems that the old card did work on a different pc but not in mine. Anyhow, I can ping either machine from the other. Now I'm working on everything else (firewall, routing, etc.). Thanks a lot to everybody (Pierre et al) who went out of their way to help me. wobo Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Correct... that was already covered in a previous post... the problem right now is to get *any* packet over the wire between the desktop/laptop... so far, nothing... And replacing a cable with a new one is no guarantee... if it was, I'd never have chased many problems where the cables were replaced 5+ times before finding one that worked... in another situation, ALL N-hundred cables were manufactured wrong... which is why one can *never* assume something is right/wrong until the problem has actually been *fixed*... :^) Pierre Richard Bown (QMW) wrote: Not so sure on that Pierre, it will get a reply from what it can see, bypassing the routing table will only get a ping back from the first device. if 100 machines are connected to the port and the port has the same address as the host then all hell breaks loose, but if the host address is different from the port address, only the port responds, because the host dos'nt know whats the otherside of the port unless its told.. ie. the routing table richard -Original Message- From: Pierre Fortin [mailto:[EMAIL PROTECTED]] Sent: Wednesday, November 28, 2001 5:22 PM To: [EMAIL PROTECTED] Subject: Re: [expert] Network problem Richard Bown (QMW) wrote: [snip] I'm waiting to hear back from Wolfgang on the list of things to verify... Until a ping -br broadcast which bypasses the routing table actually makes it to the other machine (basic connectivity), routing tables changes are not going to help (unless the network code is buggy). Cheers, Pierre Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
I think you can use that netgear card via the tulip driver. -- Asheesh. On Wed, 28 Nov 2001, Wolfgang Bornath wrote: Am Mittwoch, 28. November 2001 17:22 schrieben Sie: Richard Bown (QMW) wrote: [snip] I'm waiting to hear back from Wolfgang on the list of things to verify... Until a ping -br broadcast which bypasses the routing table actually makes it to the other machine (basic connectivity), routing tables changes are not going to help (unless the network code is buggy). That's the point. Unless I can get a connection from notebook to desktop I don't bother about routes. I ripped off the NIC which was configured as eth1 b/c it was a n old ISA card and I mistrust them. Also the MCC kept saying You only have one NIC although both showed up in the network section as up. So much to the gui tools. Right now I got a cheap NIC called Netgear FA311 Fast Ethernet PCI Card. Problem is, it comes with Linux drivers (!) but only for Red Hat 6.x and the given files will not compile b/c the makefile is counting on kgcc. Duh... So I cannot build the module. Harddrake doesn't know it, it says it's a NAtional Semiconductor Product and that it is unknown wobo (really getting in to it!) Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
On Wed, Nov 28, 2001 at 15:52 -0500, Asheesh Laroia wrote: I think you can use that netgear card via the tulip driver. -- Asheesh. No, actually it's the natsemi.o (National Semiconductor). I use it and everything works and I'm a happy camper after all :) wobo -- ... and anyway, an html can't carry a virus. (Aug 2001, Usenet) --- [EMAIL PROTECTED] ICQ: #128612867 GPG-ID: A69882EE --- ISDN4LINUX-FAQ -- Deutsch: http://www.wolf-b.de/i4l/i4lfaq-de.html Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
RE: [expert] Network problem
Ok Pierre if I remember right he said it was working with a previous version of mdk, so thats a config prob, and as he wont ack the fact that if the port has a different address to the host you have to route to everything the other sidewell just wasting my time. luckily I never moved away from 50 ohm cable linking between pc's just as well as apart from 2 ports all have radios on the end ! cya -Original Message- From: Pierre Fortin [mailto:[EMAIL PROTECTED]] Sent: Wednesday, November 28, 2001 8:48 PM To: [EMAIL PROTECTED] Subject: Re: [expert] Network problem Correct... that was already covered in a previous post... the problem right now is to get *any* packet over the wire between the desktop/laptop... so far, nothing... And replacing a cable with a new one is no guarantee... if it was, I'd never have chased many problems where the cables were replaced 5+ times before finding one that worked... in another situation, ALL N-hundred cables were manufactured wrong... which is why one can *never* assume something is right/wrong until the problem has actually been *fixed*... :^) Pierre Richard Bown (QMW) wrote: Not so sure on that Pierre, it will get a reply from what it can see, bypassing the routing table will only get a ping back from the first device. if 100 machines are connected to the port and the port has the same address as the host then all hell breaks loose, but if the host address is different from the port address, only the port responds, because the host dos'nt know whats the otherside of the port unless its told.. ie. the routing table richard -Original Message- From: Pierre Fortin [mailto:[EMAIL PROTECTED]] Sent: Wednesday, November 28, 2001 5:22 PM To: [EMAIL PROTECTED] Subject: Re: [expert] Network problem Richard Bown (QMW) wrote: [snip] I'm waiting to hear back from Wolfgang on the list of things to verify... Until a ping -br broadcast which bypasses the routing table actually makes it to the other machine (basic connectivity), routing tables changes are not going to help (unless the network code is buggy). Cheers, Pierre Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Am Dienstag, 27. November 2001 02:47 schrieb Pierre: PS: Here's one of my old .signature files: Until you've found *and* fixed a problem, you can NOT discount *any* possibility; what you gratuitously discount will likely be the source of the problem(s). Pierre Fortin - 1990 That was the result of *many* situations where people insisted it can't be X, or it *has* to be Y... By persuing the problems logically, you will generally arrive at the most likely root cause; BUT, until you've fixed it, there is still no guarantee the logical answer is the right one... Yes there is a lot of SHerlock Holmes in it ;) OK, I've spent another 3 hours, even reinstalling the whole desktop pc. It was a fresh installation anyway. I bought a new cross-over cable. I borrowed 2 other NICs from my company, Once again: Desktop has: eth0 10.0.0.10 which goes through ppp0 to the dsl modem. eth1 192.168.0.1 Io 127.0.0.1 eth0 can ping the Internet, and ping both other devices eth1 can ping both other devices both ethx can ping themselves Notebook has: eth0 192.168.0.2 Io 127.0.0.1 eth0 can ping Io and ping itself But: Notebook cannot reach any other destination apart from own eth0 and own Io Desktop cannot reach neither notebook address. BTW: Firewall is absolutely off and security is set to cracker's delight on both machines. I'm not yet bothering about internet access sharing and forewarding and masquerading, I'm just trying to get both machines to see each other. Oh, it has nothing to do that the two screens are pointing in different directions, has it? ;-)) BTW: I had everything running with a 8.o installation with the same hardware! No more ideas. Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
On Tue, 2001-11-27 at 09:12, Wolfgang Bornath wrote: OK, I've spent another 3 hours, even reinstalling the whole desktop pc. It was a fresh installation anyway. I bought a new cross-over cable. I borrowed 2 other NICs from my company, Once again: Desktop has: eth0 10.0.0.10 which goes through ppp0 to the dsl modem. eth1 192.168.0.1 Io 127.0.0.1 eth0 can ping the Internet, and ping both other devices eth1 can ping both other devices both ethx can ping themselves Notebook has: eth0 192.168.0.2 Io 127.0.0.1 eth0 can ping Io and ping itself But: Notebook cannot reach any other destination apart from own eth0 and own Io Desktop cannot reach neither notebook address. BTW: Firewall is absolutely off and security is set to cracker's delight on both machines. I'm not yet bothering about internet access sharing and forewarding and masquerading, I'm just trying to get both machines to see each other. Oh, it has nothing to do that the two screens are pointing in different directions, has it? ;-)) BTW: I had everything running with a 8.o installation with the same hardware! No more ideas. I'm running out of ideas, too. But here's a couple more: 1. I know you tested your laptop connected directly to the adsl modem and it worked before. Have you done so this time, and does it work? 2. Have you tried connecting the two PCs with a hub and CAT-5 patch cables? Dave -- When I was little, I went into a pet shop and they asked how big I'd get. -- Rodney Dangerfield msg45067/pgp0.pgp Description: PGP signature
RE: [expert] Network problem
try typing in as root route add -host 192.168.0.1 eth0 on the notebook, check the routing table and netmask on each machine. if the netmask s wrong you can get some very strange effects. you can check if this is the problem by changing the ip address of the notebook , say to 10.0.1.2 put a route in from the main pc route add -host 10.0.1.2 ethx where x is the port linked to the notebook. and on the note book route add -host 192.168.0.1 eth0 flush out any other routing just to test everything. If that works in both directions , change back to the domain of your choice and look very carefully at the netmasks on both machines. you can delete unwanted routes from the command line route del -net 192.168.0.0 netmask 255.255.0.0 eth1, thats just an example , again make sure the netmasks are correct. if you check everything out from the console command line , none of the permant routing will be lost, just what you type in , if you reboot HTH richard -Original Message- From: Wolfgang Bornath [mailto:[EMAIL PROTECTED]] Sent: Tuesday, November 27, 2001 4:13 PM To: [EMAIL PROTECTED] Subject: Re: [expert] Network problem Am Dienstag, 27. November 2001 02:47 schrieb Pierre: PS: Here's one of my old .signature files: Until you've found *and* fixed a problem, you can NOT discount *any* possibility; what you gratuitously discount will likely be the source of the problem(s). Pierre Fortin - 1990 That was the result of *many* situations where people insisted it can't be X, or it *has* to be Y... By persuing the problems logically, you will generally arrive at the most likely root cause; BUT, until you've fixed it, there is still no guarantee the logical answer is the right one... Yes there is a lot of SHerlock Holmes in it ;) OK, I've spent another 3 hours, even reinstalling the whole desktop pc. It was a fresh installation anyway. I bought a new cross-over cable. I borrowed 2 other NICs from my company, Once again: Desktop has: eth0 10.0.0.10 which goes through ppp0 to the dsl modem. eth1 192.168.0.1 Io 127.0.0.1 eth0 can ping the Internet, and ping both other devices eth1 can ping both other devices both ethx can ping themselves Notebook has: eth0 192.168.0.2 Io 127.0.0.1 eth0 can ping Io and ping itself But: Notebook cannot reach any other destination apart from own eth0 and own Io Desktop cannot reach neither notebook address. BTW: Firewall is absolutely off and security is set to cracker's delight on both machines. I'm not yet bothering about internet access sharing and forewarding and masquerading, I'm just trying to get both machines to see each other. Oh, it has nothing to do that the two screens are pointing in different directions, has it? ;-)) BTW: I had everything running with a 8.o installation with the same hardware! No more ideas. Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Wolfgang Bornath wrote: Notebook cannot reach any other destination apart from own eth0 and own Io Desktop cannot reach neither notebook address. OK... give us the results of: ping -br 255.255.255.255 ping -br 192.168.0.255 on both machines. If it works, you may something like this (from my net): $ ping -br 255.255.255.255 WARNING: pinging broadcast address PING 255.255.255.255 (255.255.255.255) from 192.168.1.100 : 56(84) bytes of data. 64 bytes from 192.168.1.100: icmp_seq=0 ttl=255 time=131 usec 64 bytes from 192.168.1.101: icmp_seq=0 ttl=255 time=525 usec (DUP!) 64 bytes from 192.168.1.102: icmp_seq=0 ttl=255 time=673 usec (DUP!) $ ping -br 192.168.1.255 WARNING: pinging broadcast address PING 192.168.1.255 (192.168.1.255) from 192.168.1.100 : 56(84) bytes of data. 64 bytes from 192.168.1.100: icmp_seq=0 ttl=255 time=134 usec 64 bytes from 192.168.1.101: icmp_seq=0 ttl=255 time=550 usec (DUP!) 64 bytes from 192.168.1.1: icmp_seq=0 ttl=150 time=554 usec (DUP!) 64 bytes from 192.168.1.102: icmp_seq=0 ttl=255 time=718 usec (DUP!) The above indicates that my LinkSys firewall (192.168.1.1) responds to net broadcasts; but not wire broadcasts. The point is that these will try to elicit a response from everyone on the wire, including self. If these pings don't work, then you'll have to figure out what is wrong physically (cable, NICs, IRQs, IOs, ...) before proceeding. If you get responses to these broadcast pings, then do: /sbin/arp -a on both machines. Should look something like this: $ /sbin/arp -a r41 (192.168.1.1) at 00:20:78:C9:E1:1A [ether] on eth0 homer (192.168.1.102) at 08:00:5A:38:6C:1B [ether] on eth0 woody (192.168.1.101) at 00:01:02:E8:D7:9F [ether] on eth0 ^ (MAC address) Make sure the hardware MAC addresses match what the other machine's ifconfig is telling you. I don't expect this to be a problem; but I'd hate to see you chase this for days only to find a bad NIC card which can only be checked from an adjacent machine (what the system *thinks* it's sending vs what actually goes out over the wire). HTH, Pierre Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Am Dienstag, 27. November 2001 15:45 schrieb Dave: On Tue, 2001-11-27 at 09:12, Wolfgang Bornath wrote: OK, I've spent another 3 hours, even reinstalling the whole desktop pc. It was a fresh installation anyway. I bought a new cross-over cable. I borrowed 2 other NICs from my company, I'm running out of ideas, too. But here's a couple more: 1. I know you tested your laptop connected directly to the adsl modem and it worked before. Have you done so this time, and does it work? Yes, and I did not change anything at the notebook, hardwareside. 2. Have you tried connecting the two PCs with a hub and CAT-5 patch cables? No, Don't have a hub, nor anyone I know of. wobo Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Am Dienstag, 27. November 2001 17:12 schrieben Sie: OK... give us the results of: ping -br 255.255.255.255 ping -br 192.168.0.255 Desktop (disconnected from internet) [root@molch wobo]# ping -br 255.255.255.255 connect: Network is unreachable ifconfig (disconnected from internet) says: eth0 Protokoll:Ethernet Hardware Adresse 00:50:BA:31:0F:B2 inet Adresse:10.0.0.10 Bcast:10.0.0.255 Maske:255.255.255.0 eth1 Protokoll:Ethernet Hardware Adresse 00:20:AF:6E:F5:16 inet Adresse:192.168.0.1 Bcast:192.168.0.255 Maske:255.255.255.0 loProtokoll:Lokale Schleife inet Adresse:127.0.0.1 Maske:255.0.0.0 # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.1 : 56(84) bytes of data. 64 bytes from 192.168.0.1: icmp_seq=0 ttl=255 time=1.844 msec 64 bytes from 192.168.0.1: icmp_seq=1 ttl=255 time=205 usec 64 bytes from 192.168.0.1: icmp_seq=2 ttl=255 time=202 usec 64 bytes from 192.168.0.1: icmp_seq=3 ttl=255 time=204 usec # ping -br 10.0.0.255 WARNING: pinging broadcast address PING 10.0.0.255 (10.0.0.255) from 10.0.0.10 : 56(84) bytes of data. 64 bytes from 10.0.0.10: icmp_seq=0 ttl=255 time=1.838 msec 64 bytes from 10.0.0.10: icmp_seq=1 ttl=255 time=188 usec 64 bytes from 10.0.0.10: icmp_seq=2 ttl=255 time=181 usec -- On the notebook : # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=92 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=73 usec # ping -br 255.255.255.255 WARNING: pinging broadcast address PING 255.255.255.255 (255.255.255.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=71 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=70 usec -- # /sbin/arp -a # arp -a gives nothing on both machines. wobo Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
hello wolfgang! I've been reading your posts about your network for a few days... great problem! that is one of these that we write on the notepad to never forget what we did :-) let's go... since arp -a gives nothing, then you have no connection, but each piece in your setup (cross-cable, notebook's nic, desktop's nic) works ok when used with the adsl (your Internet connection is trough an adsl, as I remember) or when tested with other computer I have just one opinion: by means of some power of darkness, when you connect both nics trough the cross-cable you are are shutting the electrical signals to a very low level, that isn't enough to keep the network connection. this don't happen with the adsl box... and probably wouldn't happen with a switch or a hub... why this happen? probably the notebook's nic is of poor quality... or you just run out of luck :-) orlando Wolfgang Bornath wrote: Am Dienstag, 27. November 2001 17:12 schrieben Sie: OK... give us the results of: ping -br 255.255.255.255 ping -br 192.168.0.255 Desktop (disconnected from internet) [root@molch wobo]# ping -br 255.255.255.255 connect: Network is unreachable ifconfig (disconnected from internet) says: eth0 Protokoll:Ethernet Hardware Adresse 00:50:BA:31:0F:B2 inet Adresse:10.0.0.10 Bcast:10.0.0.255 Maske:255.255.255.0 eth1 Protokoll:Ethernet Hardware Adresse 00:20:AF:6E:F5:16 inet Adresse:192.168.0.1 Bcast:192.168.0.255 Maske:255.255.255.0 loProtokoll:Lokale Schleife inet Adresse:127.0.0.1 Maske:255.0.0.0 # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.1 : 56(84) bytes of data. 64 bytes from 192.168.0.1: icmp_seq=0 ttl=255 time=1.844 msec 64 bytes from 192.168.0.1: icmp_seq=1 ttl=255 time=205 usec 64 bytes from 192.168.0.1: icmp_seq=2 ttl=255 time=202 usec 64 bytes from 192.168.0.1: icmp_seq=3 ttl=255 time=204 usec # ping -br 10.0.0.255 WARNING: pinging broadcast address PING 10.0.0.255 (10.0.0.255) from 10.0.0.10 : 56(84) bytes of data. 64 bytes from 10.0.0.10: icmp_seq=0 ttl=255 time=1.838 msec 64 bytes from 10.0.0.10: icmp_seq=1 ttl=255 time=188 usec 64 bytes from 10.0.0.10: icmp_seq=2 ttl=255 time=181 usec -- On the notebook : # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=92 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=73 usec # ping -br 255.255.255.255 WARNING: pinging broadcast address PING 255.255.255.255 (255.255.255.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=71 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=70 usec -- # /sbin/arp -a # arp -a gives nothing on both machines. wobo Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com -- ,~~v~~,,~~v~~, ,'. .', ,'. .', === + ====== + === / ~ \ / ~ \ /\_m m_/\/\_m m_/\ .\ +--+ /. / ! [EMAIL PROTECTED] ! \ / +--+ \ `\m/ \m/' `\m/ \m/' Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Am Dienstag, 27. November 2001 19:07 schrieben Sie: by means of some power of darkness, when you connect both nics trough the cross-cable you are are shutting the electrical signals to a very low level, that isn't enough to keep the network connection. this don't happen with the adsl box... and probably wouldn't happen with a switch or a hub... why this happen? probably the notebook's nic is of poor quality... or you just run out of luck :-) Sounds good, BUT I think I mentioned that I had the same hardware running (notebook to desktop via c/o cable) before with 8.0. So it HAS to be a problem of setup, or not? wobo Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Wolfgang Bornath wrote: Am Dienstag, 27. November 2001 17:12 schrieben Sie: OK... give us the results of: ping -br 255.255.255.255 ping -br 192.168.0.255 Desktop (disconnected from internet) [root@molch wobo]# ping -br 255.255.255.255 connect: Network is unreachable ifconfig (disconnected from internet) says: eth0 Protokoll:Ethernet Hardware Adresse 00:50:BA:31:0F:B2 inet Adresse:10.0.0.10 Bcast:10.0.0.255 Maske:255.255.255.0 eth1 Protokoll:Ethernet Hardware Adresse 00:20:AF:6E:F5:16 inet Adresse:192.168.0.1 Bcast:192.168.0.255 Maske:255.255.255.0 loProtokoll:Lokale Schleife inet Adresse:127.0.0.1 Maske:255.0.0.0 # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.1 : 56(84) bytes of data. 64 bytes from 192.168.0.1: icmp_seq=0 ttl=255 time=1.844 msec 64 bytes from 192.168.0.1: icmp_seq=1 ttl=255 time=205 usec 64 bytes from 192.168.0.1: icmp_seq=2 ttl=255 time=202 usec 64 bytes from 192.168.0.1: icmp_seq=3 ttl=255 time=204 usec # ping -br 10.0.0.255 WARNING: pinging broadcast address PING 10.0.0.255 (10.0.0.255) from 10.0.0.10 : 56(84) bytes of data. 64 bytes from 10.0.0.10: icmp_seq=0 ttl=255 time=1.838 msec 64 bytes from 10.0.0.10: icmp_seq=1 ttl=255 time=188 usec 64 bytes from 10.0.0.10: icmp_seq=2 ttl=255 time=181 usec -- On the notebook : # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=92 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=73 usec # ping -br 255.255.255.255 WARNING: pinging broadcast address PING 255.255.255.255 (255.255.255.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=71 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=70 usec -- # /sbin/arp -a # arp -a gives nothing on both machines. The lack of (DUP!) responses confirms that neither machine is seeing the other's packets... like it or not, you appear to have a physical problem... ping -r bypasses the routing table and sends directly to the NIC... Next steps (assumes your NICs have link traffic LEDs): 1. see if your link LEDs are on on both NICs 2. make sure you have the right drivers installed 3. start pinging (broadcasts as above) from one machine at a time and see if the traffic LED flashes on the card itself... 4. make sure the packet gets to the other end of the cable by checking traffic LED at other machine Until you get past this point, I'm not sure anything else will help... actually, now that I think about it... you could try to force the use of unicast packets only (avoiding the ARP broadcast in the case of directed pings) by setting the arp entries manually (see man arp) in both machines so their arp tables have the IP-to-h/w addresses pre-resolved. BTW, ping -I eth1 -br 255.255.255.255 should overcome the Network is unreachable response... if so, ping may have a bug 'cuz -r is supposed to bypass the routing table. Of course, it could be that by doing so, it assumes 255.* is a network and tries to find the corresponding interface... oh well... not the first anomoly I've found with ping... Pierre Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Am Dienstag, 27. November 2001 17:12 schrieben Sie: OK... give us the results of: ping -br 255.255.255.255 ping -br 192.168.0.255 Desktop (disconnected from internet) [root@molch wobo]# ping -br 255.255.255.255 connect: Network is unreachable ifconfig (disconnected from internet) says: eth0 Protokoll:Ethernet Hardware Adresse 00:50:BA:31:0F:B2 inet Adresse:10.0.0.10 Bcast:10.0.0.255 Maske:255.255.255.0 eth1 Protokoll:Ethernet Hardware Adresse 00:20:AF:6E:F5:16 inet Adresse:192.168.0.1 Bcast:192.168.0.255 Maske:255.255.255.0 loProtokoll:Lokale Schleife inet Adresse:127.0.0.1 Maske:255.0.0.0 # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.1 : 56(84) bytes of data. 64 bytes from 192.168.0.1: icmp_seq=0 ttl=255 time=1.844 msec 64 bytes from 192.168.0.1: icmp_seq=1 ttl=255 time=205 usec 64 bytes from 192.168.0.1: icmp_seq=2 ttl=255 time=202 usec 64 bytes from 192.168.0.1: icmp_seq=3 ttl=255 time=204 usec # ping -br 10.0.0.255 WARNING: pinging broadcast address PING 10.0.0.255 (10.0.0.255) from 10.0.0.10 : 56(84) bytes of data. 64 bytes from 10.0.0.10: icmp_seq=0 ttl=255 time=1.838 msec 64 bytes from 10.0.0.10: icmp_seq=1 ttl=255 time=188 usec 64 bytes from 10.0.0.10: icmp_seq=2 ttl=255 time=181 usec -- On the notebook : # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=92 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=73 usec # ping -br 255.255.255.255 WARNING: pinging broadcast address PING 255.255.255.255 (255.255.255.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=71 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=70 usec -- # /sbin/arp -a # arp -a gives nothing on both machines. wobo Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
RE: [expert] Network problem
Hi all, It would help Wolfgang if you could send the routing table on both machines ! It dos'nt matter how good your sigs are at each end of the cable if a return route is incorrect. Its only a point to point system there, the fun starts when you use 6 interfaces on the same machine all with different ip addresses and all routing different subnets. But at least I have the advantages of a class A ip address which there are no duplicates anywhere, and I have control over 4 subnets -Original Message- From: Pierre Fortin [mailto:[EMAIL PROTECTED]] Sent: Tuesday, November 27, 2001 7:19 PM To: [EMAIL PROTECTED] Subject: Re: [expert] Network problem Wolfgang Bornath wrote: Am Dienstag, 27. November 2001 17:12 schrieben Sie: OK... give us the results of: ping -br 255.255.255.255 ping -br 192.168.0.255 Desktop (disconnected from internet) [root@molch wobo]# ping -br 255.255.255.255 connect: Network is unreachable ifconfig (disconnected from internet) says: eth0 Protokoll:Ethernet Hardware Adresse 00:50:BA:31:0F:B2 inet Adresse:10.0.0.10 Bcast:10.0.0.255 Maske:255.255.255.0 eth1 Protokoll:Ethernet Hardware Adresse 00:20:AF:6E:F5:16 inet Adresse:192.168.0.1 Bcast:192.168.0.255 Maske:255.255.255.0 loProtokoll:Lokale Schleife inet Adresse:127.0.0.1 Maske:255.0.0.0 # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.1 : 56(84) bytes of data. 64 bytes from 192.168.0.1: icmp_seq=0 ttl=255 time=1.844 msec 64 bytes from 192.168.0.1: icmp_seq=1 ttl=255 time=205 usec 64 bytes from 192.168.0.1: icmp_seq=2 ttl=255 time=202 usec 64 bytes from 192.168.0.1: icmp_seq=3 ttl=255 time=204 usec # ping -br 10.0.0.255 WARNING: pinging broadcast address PING 10.0.0.255 (10.0.0.255) from 10.0.0.10 : 56(84) bytes of data. 64 bytes from 10.0.0.10: icmp_seq=0 ttl=255 time=1.838 msec 64 bytes from 10.0.0.10: icmp_seq=1 ttl=255 time=188 usec 64 bytes from 10.0.0.10: icmp_seq=2 ttl=255 time=181 usec -- On the notebook : # ping -br 192.168.0.255 WARNING: pinging broadcast address PING 192.168.0.255 (192.168.0.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=92 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=73 usec # ping -br 255.255.255.255 WARNING: pinging broadcast address PING 255.255.255.255 (255.255.255.255) from 192.168.0.2 : 56(84) bytes of data. 64 bytes from 192.168.0.2: icmp_seq=0 ttl=255 time=71 msec 64 bytes from 192.168.0.2: icmp_seq=1 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=2 ttl=255 time=73 usec 64 bytes from 192.168.0.2: icmp_seq=3 ttl=255 time=70 usec -- # /sbin/arp -a # arp -a gives nothing on both machines. The lack of (DUP!) responses confirms that neither machine is seeing the other's packets... like it or not, you appear to have a physical problem... ping -r bypasses the routing table and sends directly to the NIC... Next steps (assumes your NICs have link traffic LEDs): 1. see if your link LEDs are on on both NICs 2. make sure you have the right drivers installed 3. start pinging (broadcasts as above) from one machine at a time and see if the traffic LED flashes on the card itself... 4. make sure the packet gets to the other end of the cable by checking traffic LED at other machine Until you get past this point, I'm not sure anything else will help... actually, now that I think about it... you could try to force the use of unicast packets only (avoiding the ARP broadcast in the case of directed pings) by setting the arp entries manually (see man arp) in both machines so their arp tables have the IP-to-h/w addresses pre-resolved. BTW, ping -I eth1 -br 255.255.255.255 should overcome the Network is unreachable response... if so, ping may have a bug 'cuz -r is supposed to bypass the routing table. Of course, it could be that by doing so, it assumes 255.* is a network and tries to find the corresponding interface... oh well... not the first anomoly I've found with ping... Pierre Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Dave said: As long as your desktop is connecting to the Internet ok, then we can probably assume everything on that side is setup correctly. OK I think the next thing I would try to do is traceroute from your laptop, to some IP address on the internet. That way, we can see just where your TCP packets are getting hung up. First I checked that the network card in the notebook is ok. I connected the line from the ADSL modem to the notebook, configured the internet access and was online. So the NIC works. I don't get anywhere from the notebook as long as it is connected to the desktop. I looked at the routing table and once more checked all the IPs. No way. Right now I'm writing this from the office so I don't have access to my box until 1600 local time (GMT +1). -- wobo http://www.wolf-b.de ISDN4LINUX FAQ - DE ICQ #128612867 The source is out there! Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
hi just a long shot ! check /etc/sysconfig/network is forwarding set to either yes or true ?... it should be its very easy to miss regards richard On Mon, 2001-11-26 at 04:17, Dave Sherman wrote: On Sun, 2001-11-25 at 13:42, Wolfgang Bornath wrote: Am Sonntag, 25. November 2001 17:36 schrieben Sie: Couple of questions for you: 1. Can you ping your own interfaces? That is, from your desktop can you ping each of the cards' IP addresses? Can you ping 127.0.0.1? Can you also ping your laptop's card from itself (the real IP and 127.0.0.1)? If you can't do that, then you obviously won't be going anywhere outside... Yes to all, all NICs can ping their own address and the NICs in each machine can ping the respective Io addresses. 2. Do you have a firewall running on either of the PCs? Sometimes Tried that road already, no success. 3. Do you have internet connection sharing enabled in the desktop PC? If not, it won't be able to forward packets from your laptop out to the internet. Yes I did. 4. Can your desktop reach the internet? You seem to imply that you can, but I am not sure. Yes it can 5. Am I understanding that your desktop's external interface (eth1) has two IP addresses? Can you give more information about your ADSL The external interface is eth0. It is linked to ppp0 which presents itself as the internet connector. ADSL in Germany (if you don't have a router) works like this: From the wall socket your telephone line (ISDN) goes to a splitter. From there one line goes to the ISDN box and further to the ISDN devices. The other line goes to the DSL modem and from there to the NIC in the pc. Software used is the pppoe package. I will try to give more help when I see your answers. I appreciate the time you waste in this. As long as your desktop is connecting to the Internet ok, then we can probably assume everything on that side is setup correctly. I think the next thing I would try to do is traceroute from your laptop, to some IP address on the internet. That way, we can see just where your TCP packets are getting hung up. To traceroute: $ /usr/sbin/traceroute 208.20.203.226 The IP address is the static IP of my DSL router. Feel free to try mine, or another address closer to you. Dave Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Wolfgang Bornath wrote: Hi, I have a desktop and a notebook, both connected via cross-over cable. All in all it looks ok to me. But the desktop doesn't ping the notebook and vice versa. The notebook can't connect to the internet. The cable was the last resort but it's ok, I tested it at a friend's machine who has a similar setup. What is wrong here? Your cable (or flaky connections)... Note that NEITHER box is seeing any traffic from the other: RX packets:0 Pierre Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Am Montag, 26. November 2001 18:37 schrieben Sie: Wolfgang Bornath wrote: Hi, I have a desktop and a notebook, both connected via cross-over cable. All in all it looks ok to me. But the desktop doesn't ping the notebook and vice versa. The notebook can't connect to the internet. The cable was the last resort but it's ok, I tested it at a friend's machine who has a similar setup. What is wrong here? Your cable (or flaky connections)... I already tested that at a friend's setup, s.a. I also tested the NICs singularly, they are doing ok. wobo Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Dave said: Did you try the traceroute command I gave you? Here it is again: /usr/sbin/traceroute 208.20.203.226 Run this in an xterm and watch the results. When your packets stop As I expected they aren't doing the first hop of the dance. At this point, I am suspecting the cable, in spite of the fact that you tested it at your friend's house. But I could certainly be wrong. OK, I'll switch cable and try to get hold of another NIC to test. Thanks for trying, all of you. -- wobo http://www.wolf-b.de ISDN4LINUX FAQ - DE ICQ #128612867 The source is out there! Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
Wolfgang Bornath wrote: Am Montag, 26. November 2001 18:37 schrieben Sie: Wolfgang Bornath wrote: Hi, I have a desktop and a notebook, both connected via cross-over cable. All in all it looks ok to me. But the desktop doesn't ping the notebook and vice versa. The notebook can't connect to the internet. The cable was the last resort but it's ok, I tested it at a friend's machine who has a similar setup. What is wrong here? Your cable (or flaky connections)... I already tested that at a friend's setup, s.a. I also tested the NICs singularly, they are doing ok. Then, the first thing to resolve is why you are not seeing any packets coming in on either side of that cable... For example, pinging the desktop from the laptop should show the TX packets incrementing on the laptop and RX packets incrementing on the desktop... if not, why not. Pierre PS: Here's one of my old .signature files: Until you've found *and* fixed a problem, you can NOT discount *any* possibility; what you gratuitously discount will likely be the source of the problem(s). Pierre Fortin - 1990 That was the result of *many* situations where people insisted it can't be X, or it *has* to be Y... By persuing the problems logically, you will generally arrive at the most likely root cause; BUT, until you've fixed it, there is still no guarantee the logical answer is the right one... Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
On Sat, 2001-11-24 at 20:38, Wolfgang Bornath wrote: Hi, I have a desktop and a notebook, both connected via cross-over cable. The desktop is connected to internet via ADSL. Both machines run MDK 8.1 Desktop: eth0 RealTek RTL-8029 module = ne2k-pci IRQ 12 eth1 3com 3C509 module = 3c509 IRQ10 (ISA) Both modules load at bootup. ifconfig shows (while eth0 is connected as ppp0 to the internet): eth0 Protokoll:Ethernet Hardware Adresse 00:50:BA:31:0F:B2 inet Adresse:10.0.0.10 Bcast:10.0.0.255 Maske:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:93 errors:0 dropped:0 overruns:0 frame:0 TX packets:96 errors:0 dropped:0 overruns:0 carrier:0 Kollisionen:0 Sendewarteschlangenlänge:100 RX bytes:65325 (63.7 Kb) TX bytes:14185 (13.8 Kb) Interrupt:12 Basisadresse:0xe800 eth1 Protokoll:Ethernet Hardware Adresse 00:20:AF:6E:F5:16 inet Adresse:192.168.0.1 Bcast:192.168.0.255 Maske:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:16 errors:0 dropped:0 overruns:0 carrier:1 Kollisionen:0 Sendewarteschlangenlänge:100 RX bytes:0 (0.0 b) TX bytes:672 (672.0 b) Interrupt:10 Basisadresse:0x300 route -n shows: 217.5.98.14 0.0.0.0 255.255.255.255 UH 0 0 0 ppp0 10.0.0.00.0.0.0 255.255.255.0 U 0 0 0 eth0 192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0 eth1 127.0.0.0 0.0.0.0 255.0.0.0 U 0 0 0 lo 0.0.0.0 217.5.98.140.0.0.0 UG 0 0 0 ppp0 -- Notebook: eth0 SIS900 module = sis900 IRQ 10 Module loads ifconfig says: eth0 Protokoll:Ethernet Hardware Adresse 00:a0:cc:c6:b3:58 inet Adresse:192.168.0.2 Bcast:192.168.0.255 Maske:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:44 errors:0 dropped:0 overruns:0 carrier:0 Kollisionen:0 Sendewarteschlangenlänge:100 RX bytes: 0 (0.0 Kb) TX bytes:3615 (3.5 Kb) Interrupt:10 Basisadresse:0xd000 route -n says: 192.168.0.0 0.0.0.0 255.255.255.0 U 0 0 0 eth0 127.0.0.0 0.0.0.0 255.0.0.0U 0 0 0 lo 0.0.0.0 192.168.0.10.0.0.0 UG 0 0 0 eth0 All in all it looks ok to me. But the desktop doesn't ping the notebook and vice versa. The notebook can't connect to the internet. The cable was the last resort but it's ok, I tested it at a friend's machine who has a similar setup. What is wrong here? wobo Couple of questions for you: 1. Can you ping your own interfaces? That is, from your desktop can you ping each of the cards' IP addresses? Can you ping 127.0.0.1? Can you also ping your laptop's card from itself (the real IP and 127.0.0.1)? If you can't do that, then you obviously won't be going anywhere outside... 2. Do you have a firewall running on either of the PCs? Sometimes firewalls will block ICMP (ping) packets, and so your pings are unable to work. Try turning off your firewall -- in MDK 8.1, assuming you are using the built-in firewall, run (as root) '/etc/init.d/iptables stop' but without the quotes. This is for a 2.4 kernel. If you are using the 2.2 kernel, try '/etc/init.d/ipchains stop'. If you are not using the built-in firewall, then just try (as root) 'iptables -F' or 'ipchains -F' -- these will flush the firewall rules and allow anything. Just restart your firewall to bring the rules back. 3. Do you have internet connection sharing enabled in the desktop PC? If not, it won't be able to forward packets from your laptop out to the internet. 4. Can your desktop reach the internet? You seem to imply that you can, but I am not sure. 5. Am I understanding that your desktop's external interface (eth1) has two IP addresses? Can you give more information about your ADSL setup? I am only familiar with the kind that uses a router or bridge to connect, so that the router/bridge gets a phone wire from the wall, and has an ethernet wire going out to eihter a hub or a single PC with a standard NIC. This device, if a router, usually runs a dhcp server and provides all connection information needed for the PCs. If it's a bridge, then the PCs either need to be running dhcp and get their info from the ISP, or they need to have static IP addresses (and other network stuff, like DNS and default gateway) assigned. I will try to give more help when I see your answers. Dave -- You may worry about your hair-do today, but tomorrow much peanut butter will be sold. msg44900/pgp0.pgp Description: PGP signature
Re: [expert] Network problem
Am Sonntag, 25. November 2001 17:36 schrieben Sie: Couple of questions for you: 1. Can you ping your own interfaces? That is, from your desktop can you ping each of the cards' IP addresses? Can you ping 127.0.0.1? Can you also ping your laptop's card from itself (the real IP and 127.0.0.1)? If you can't do that, then you obviously won't be going anywhere outside... Yes to all, all NICs can ping their own address and the NICs in each machine can ping the respective Io addresses. 2. Do you have a firewall running on either of the PCs? Sometimes Tried that road already, no success. 3. Do you have internet connection sharing enabled in the desktop PC? If not, it won't be able to forward packets from your laptop out to the internet. Yes I did. 4. Can your desktop reach the internet? You seem to imply that you can, but I am not sure. Yes it can 5. Am I understanding that your desktop's external interface (eth1) has two IP addresses? Can you give more information about your ADSL The external interface is eth0. It is linked to ppp0 which presents itself as the internet connector. ADSL in Germany (if you don't have a router) works like this: From the wall socket your telephone line (ISDN) goes to a splitter. From there one line goes to the ISDN box and further to the ISDN devices. The other line goes to the DSL modem and from there to the NIC in the pc. Software used is the pppoe package. I will try to give more help when I see your answers. I appreciate the time you waste in this. wobo Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com
Re: [expert] Network problem
On Sun, 2001-11-25 at 13:42, Wolfgang Bornath wrote: Am Sonntag, 25. November 2001 17:36 schrieben Sie: Couple of questions for you: 1. Can you ping your own interfaces? That is, from your desktop can you ping each of the cards' IP addresses? Can you ping 127.0.0.1? Can you also ping your laptop's card from itself (the real IP and 127.0.0.1)? If you can't do that, then you obviously won't be going anywhere outside... Yes to all, all NICs can ping their own address and the NICs in each machine can ping the respective Io addresses. 2. Do you have a firewall running on either of the PCs? Sometimes Tried that road already, no success. 3. Do you have internet connection sharing enabled in the desktop PC? If not, it won't be able to forward packets from your laptop out to the internet. Yes I did. 4. Can your desktop reach the internet? You seem to imply that you can, but I am not sure. Yes it can 5. Am I understanding that your desktop's external interface (eth1) has two IP addresses? Can you give more information about your ADSL The external interface is eth0. It is linked to ppp0 which presents itself as the internet connector. ADSL in Germany (if you don't have a router) works like this: From the wall socket your telephone line (ISDN) goes to a splitter. From there one line goes to the ISDN box and further to the ISDN devices. The other line goes to the DSL modem and from there to the NIC in the pc. Software used is the pppoe package. I will try to give more help when I see your answers. I appreciate the time you waste in this. As long as your desktop is connecting to the Internet ok, then we can probably assume everything on that side is setup correctly. I think the next thing I would try to do is traceroute from your laptop, to some IP address on the internet. That way, we can see just where your TCP packets are getting hung up. To traceroute: $ /usr/sbin/traceroute 208.20.203.226 The IP address is the static IP of my DSL router. Feel free to try mine, or another address closer to you. Dave -- Hindsight is always 20:20. -- Billy Wilder msg44932/pgp0.pgp Description: PGP signature
Re: [expert] Network problem: problem with 3Com card setup? [Was:can't connect via DHCP]
David, Thank you very much for your help, it took me a while to digect it, plus other things changed as well. Most importantly -- DHCP is not the source of my problem. Our computer admins assigned me a static IP address (with a subnet mask, gateway and 3 DNS servers), but the story is the same -- it works under Windows (I plugged the IP numbers instead of DHCP, just to see that it works in Win), but not in Linux... I've done networking setup many times since, through linuxconf and directly editing /etc/hosts, /etc/resolv.conf, etc... I start to think that there is a problem with the card setup. Maybe a bad IRQ sharing? Is there any way to run some diagnostic in Linux to see if the card setup is ok? I am grateful for any suggestions... David Rankin wrote: Uhh.., I think I know what your problem is. (Guys on the list -- tell me if I'm wrong) Your modules files should be '.o' not '.o.gz'. I don't think modprobe or anything else for that reason will unzip a module for use. Here is how to test the 3c90x.o driver. No, it seems it is the other way around -- modprobe seems to look for a .o.gz file. I uncompressed the 3c90x module (btw, all modules in drivers are compressed), and then modprobe couldn't find it: [root@botik net]# pwd /lib/modules/2.4.3-20mdk/kernel/drivers/net [root@botik net]# gunzip 3c90x.o.gz [root@botik net]# cd .. [root@botik drivers]# ll net/3c* -rw-r--r--1 root root 4078 Apr 15 17:14 net/3c501.o.gz -rw-r--r--1 root root 4006 Apr 15 17:14 net/3c503.o.gz -rw-r--r--1 root root 9326 Apr 15 17:14 net/3c505.o.gz -rw-r--r--1 root root 4641 Apr 15 17:14 net/3c507.o.gz -rw-r--r--1 root root 5890 Apr 15 17:14 net/3c509.o.gz -rw-r--r--1 root root 9910 Apr 15 17:14 net/3c515.o.gz -rw-r--r--1 root root16191 Apr 15 17:14 net/3c59x.o.gz -rw-r--r--1 root root31276 Apr 15 17:14 net/3c90x.o [root@botik drivers]# modprobe -t net 3c90x modprobe: Can't locate module 3c90x which is needed for 3c90x Then I compressed it back, and it's found, but doesn't load: [root@botik drivers]# ll net/3c* -rw-r--r--1 root root 4078 Apr 15 17:14 net/3c501.o.gz -rw-r--r--1 root root 4006 Apr 15 17:14 net/3c503.o.gz -rw-r--r--1 root root 9326 Apr 15 17:14 net/3c505.o.gz -rw-r--r--1 root root 4641 Apr 15 17:14 net/3c507.o.gz -rw-r--r--1 root root 5890 Apr 15 17:14 net/3c509.o.gz -rw-r--r--1 root root 9910 Apr 15 17:14 net/3c515.o.gz -rw-r--r--1 root root16191 Apr 15 17:14 net/3c59x.o.gz -rw-r--r--1 root root14468 Apr 15 17:14 net/3c90x.o.gz [root@botik drivers]# modprobe -t net 3c90x Warning: /lib/modules/2.4.3-20mdk/kernel/drivers/net/3c90x.o.gz parameter switchdelay has max min! /lib/modules/2.4.3-20mdk/kernel/drivers/net/3c90x.o.gz: init_module: No such device Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters /lib/modules/2.4.3-20mdk/kernel/drivers/net/3c90x.o.gz: insmod /lib/modules/2.4.3-20mdk/kernel/drivers/net/3c90x.o.gz failed /lib/modules/2.4.3-20mdk/kernel/drivers/net/3c90x.o.gz: insmod 3c90x failed (1) change to your modules directory for your system. Mine is /lib/modules/2.2.17-21mdk (2) the network modules should be in /lib/modules/2.2.17-21mdk/net It seems Mandrake 8.0 uses a slightly different layout: I found 3c90x.o.gz in /lib/modules/2.4.3-20mdk/kernel/drivers/net (see above). (3) from /lib/modules/2.2.17-21mdk do ls net/3c* to make sure you have 3c90x.o (4) do modprobe -t net 3c90x or modprobe -t net 3c* (to try all 3com drivers) Did that (see above), 3c59x loads just fine, 3c90x fails... (5) after doing modprobe, do lsmod. If modprobe was able to load the driver it will be listed lsmod lists 3c59x, seems ok, but Ethernet connection doesn't work... [root@botik drivers]# lsmod Module Size Used by es1371 27440 0 soundcore 3504 4 [es1371] ac97_codec 8688 0 [es1371] gameport1520 0 [es1371] 3c59x 24640 1 (autoclean) usb-uhci 20672 0 (unused) usbcore47248 1 [usb-uhci] ide-scsi7568 0 nls_iso8859-1 2848 1 (autoclean) nls_cp850 3584 1 (autoclean) vfat9040 1 (autoclean) fat30720 0 (autoclean) [vfat] supermount 32496 8 (autoclean) sd_mod 11048 0 (unused) scsi_mod 86036 2 [ide-scsi sd_mod] Thank you again, Aleksey
Re: [expert] Network problem: can't connect via DHCP
Aleksey Naumov wrote: David, thank you for your help. I still can't resolve my problem (see below)... David Rankin wrote: Aleksey Naumov wrote: Hi, Mandrake experts! I've just installed Mandrake 8.0 on my PII 400MHz HP at work. My computer is on a college network and connection is done with DHCP. The problem is that I can't get DHCP connection to work in Linux, while it works just fine in Win98. At boot time I get: Setting network parameters:[ OK ] Bringing up interface lo: [ OK ] Bringing up interface eth0: Determining IP information for eth0... failed. [FAILED] Could this be a problem with my Ethernet card rather than DHCP? It doesn't look like it to me (especially since it works perfectly under Win98), but just in case, here the info on my Ethernet card: Yes this is probably a problem with your ethernet card and module configuration I don't know how to check if the card itself works, but the following messages (from dmesg) make me think it's ok: read the man page on modprobe and then test the 3c905 module with modprobe -t Obviously, the direct approach with insmod 3c90x gives the same messages... Where do I go from here? How can I check if my problem is related to module configuration or not? Uhh.., I think I know what your problem is. (Guys on the list -- tell me if I'm wrong) Your modules files should be '.o' not '.o.gz'. I don't think modprobe or anything else for that reason will unzip a module for use. Here is how to test the 3c90x.o driver. (1) change to your modules directory for your system. Mine is /lib/modules/2.2.17-21mdk (2) the network modules should be in /lib/modules/2.2.17-21mdk/net (3) from /lib/modules/2.2.17-21mdk do ls net/3c* to make sure you have 3c90x.o (4) do modprobe -t net 3c90x or modprobe -t net 3c* (to try all 3com drivers) (5) after doing modprobe, do lsmod. If modprobe was able to load the driver it will be listed (6) restart networking /etc/rc.d/init.d/network restart (7) if it works add 'alias eth0 3c90x' (or whatever driver worked) to /etc/modules.conf (8) if it didn't work - go to step 1 and try a different driver. I apologize if my recollection of the syntax is not 100% correct (I'm at home on the Windoze box) and I'm getting old (past 30) so my memory is beginning to go. -- David Rankin Nacogdoches, Texas
Re: [expert] Network problem: can't connect via DHCP
I'm on a similar campus, but DHCP works fine for me :-D But that doesn't help you out ; ) I suggest taking the things your Help Desk people told you about (subnet mask, IP address, etc.) and hard-coding them rather than trying to get DHCP to automatically configure your settings for you. Offhand I'm not sure what graphical utility does this, but I'm sure it's either the main Control Center or something similar. All you have to do is click on your ethernet card and then manually set these values. This *should* work. Good luck Sincerely, Eric Eric Krout Email: [EMAIL PROTECTED] Web: http://www.eg.bucknell.edu/~ekrout (Soon, EricKrout.com) Bucknell Computer Science Engineering '03 Chairman, Bucknell's Assoc. for Computing Machinery (ACM) On Fri, 27 Apr 2001, Aleksey Naumov wrote: Hi, Mandrake experts! I've just installed Mandrake 8.0 on my PII 400MHz HP at work. My computer is on a college network and connection is done with DHCP. The problem is that I can't get DHCP connection to work in Linux, while it works just fine in Win98. At boot time I get: Setting network parameters:[ OK ] Bringing up interface lo: [ OK ] Bringing up interface eth0: Determining IP information for eth0... failed. [FAILED] I've tried to play with dhcpcd parameters, but with no result: dhcpcd -v DHCP Client Daemon v.1.3.19-pl4 dhcpcd dhcpcd -r as well as playing woth the -h (host) option: dhcpcd -h foohost dhcpcd -h machine_name # name my machine is known as under Win dhcpcd -h machine_name.domainname I get the same message (in log): Apr 27 19:41:41 HAP429 dhcpcd[1355]: broadcasting DHCP_DISCOVER Apr 27 19:42:41 HAP429 dhcpcd[1355]: timed out waiting for a valid DHCP server response I also tried to use the pump daemon instead of dhcpcd: pump Operation failed. pump --win-client-ident Operation failed. as well as all -h options, but with the same result. At least pump daemon logs more info, but I don't know if it's useful. Does anyone? Here it is: Apr 27 19:37:23 HAP429 pumpd[1348]: PUMP: sending discover Apr 27 19:37:23 HAP429 pumpd[1348]: breq: opcode: 1 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: hw: 1 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: hwlength: 6 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: hopcount: 0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: xid: 0xf5fba69c Apr 27 19:37:23 HAP429 pumpd[1348]: breq: secs: 0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: flags: 0x Apr 27 19:37:23 HAP429 pumpd[1348]: breq: ciaddr: 0.0.0.0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: yiaddr: 0.0.0.0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: server_ip: 0.0.0.0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: bootp_gw_ip: 0.0.0.0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: hwaddr: Apr 27 19:37:23 HAP429 pumpd[1348]: breq: servername: Apr 27 19:37:23 HAP429 pumpd[1348]: breq: bootfile: Apr 27 19:37:23 HAP429 pumpd[1348]: breq: vendor: 0x63 0x53 0x82 0x63 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: vendor: 53 1 0x01 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: vendor: 0xff Apr 27 19:37:23 HAP429 kernel: eth0: using NWAY device table, not 8 People in tech support were not very helpful (it's a Windows only campus). They gave me IP addresses for subnet, gateway and DHCP server, but what do I do with them? The DHCP server is probably run on NT. I've read the DHCP mini-howto, as well as other network howtos and I can't get this moving. I also went through network setup in Mandrake control center (as well as linuxconf), the Mandrake control center uses wizard and determines correctly that I am on a LAN and using DHCP, then reports that the network is set... and it isn't! Any ideas would be very much appreciated! Aleksey P.S. Could this be a problem with my Ethernet card rather than DHCP? It doesn't look like it to me (especially since it works perfectly under Win98), but just in case, here the info on my Ethernet card: - I have a 3Com 3c905B-Combo [Deluxe EtherLink XL 10/100] card. Of the 3 connectors the one that's is used is 10Base2 (thinnet with metal push-and-turn-to-lock). The driver used is 3c59x that came with the system. I tried to compile the newer 3c90x, but couldn't get it to compile. However, this is likely not a problem, since the Ethernet-HOWTO says that my card (3c905B) is supported by this driver (3c59x). I don't know how to check if the card itself works, but the following messages (from dmesg) make me think it's ok: dmesg ... 3c59x.c:LK1.1.13 27 Jan 2001 Donald Becker and others. http://www.scyld.com/network/vortex.html See Documentation/networking/vortex.txt eth0: 3Com PCI 3c905B Cyclone 10/100/BNC at 0x1400, 00:50:04:0a:fb:13, IRQ 11 product code 5545 rev 00.0 date 01-21-99 8K
Re: [expert] Network problem: can't connect via DHCP
When it rains, it pours. Someone else just asked the exact same question. Ive been through it myself, and so here's what i said: When i experienced that symptom (timeout), it was because my ISP required that i pass a hostname with the request: dhcpcd -h cr843732-a was the thing it needed. To make that an automatic phenomenon during bootup, make sure to include the following in /etc/sysconfig/network-scripts/ifcfg-eth0 (for eth0): NEED_HOSTNAME=1 DHCP_HOSTNAME=cr843732-a So check that out, it might be it. j --- Aleksey Naumov [EMAIL PROTECTED] wrote: Hi, Mandrake experts! I've just installed Mandrake 8.0 on my PII 400MHz HP at work. My computer is on a college network and connection is done with DHCP. The problem is that I can't get DHCP connection to work in Linux, while it works just fine in Win98. At boot time I get: Setting network parameters:[ OK ] Bringing up interface lo: [ OK ] Bringing up interface eth0: Determining IP information for eth0... failed. [FAILED] I've tried to play with dhcpcd parameters, but with no result: dhcpcd -v DHCP Client Daemon v.1.3.19-pl4 dhcpcd dhcpcd -r as well as playing woth the -h (host) option: dhcpcd -h foohost dhcpcd -h machine_name # name my machine is known as under Win dhcpcd -h machine_name.domainname I get the same message (in log): Apr 27 19:41:41 HAP429 dhcpcd[1355]: broadcasting DHCP_DISCOVER Apr 27 19:42:41 HAP429 dhcpcd[1355]: timed out waiting for a valid DHCP server response I also tried to use the pump daemon instead of dhcpcd: pump Operation failed. pump --win-client-ident Operation failed. as well as all -h options, but with the same result. At least pump daemon logs more info, but I don't know if it's useful. Does anyone? Here it is: Apr 27 19:37:23 HAP429 pumpd[1348]: PUMP: sending discover Apr 27 19:37:23 HAP429 pumpd[1348]: breq: opcode: 1 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: hw: 1 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: hwlength: 6 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: hopcount: 0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: xid: 0xf5fba69c Apr 27 19:37:23 HAP429 pumpd[1348]: breq: secs: 0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: flags: 0x Apr 27 19:37:23 HAP429 pumpd[1348]: breq: ciaddr: 0.0.0.0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: yiaddr: 0.0.0.0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: server_ip: 0.0.0.0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: bootp_gw_ip: 0.0.0.0 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: hwaddr: Apr 27 19:37:23 HAP429 pumpd[1348]: breq: servername: Apr 27 19:37:23 HAP429 pumpd[1348]: breq: bootfile: Apr 27 19:37:23 HAP429 pumpd[1348]: breq: vendor: 0x63 0x53 0x82 0x63 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: vendor: 53 1 0x01 Apr 27 19:37:23 HAP429 pumpd[1348]: breq: vendor: 0xff Apr 27 19:37:23 HAP429 kernel: eth0: using NWAY device table, not 8 People in tech support were not very helpful (it's a Windows only campus). They gave me IP addresses for subnet, gateway and DHCP server, but what do I do with them? The DHCP server is probably run on NT. I've read the DHCP mini-howto, as well as other network howtos and I can't get this moving. I also went through network setup in Mandrake control center (as well as linuxconf), the Mandrake control center uses wizard and determines correctly that I am on a LAN and using DHCP, then reports that the network is set... and it isn't! Any ideas would be very much appreciated! Aleksey P.S. Could this be a problem with my Ethernet card rather than DHCP? It doesn't look like it to me (especially since it works perfectly under Win98), but just in case, here the info on my Ethernet card: - I have a 3Com 3c905B-Combo [Deluxe EtherLink XL 10/100] card. Of the 3 connectors the one that's is used is 10Base2 (thinnet with metal push-and-turn-to-lock). The driver used is 3c59x that came with the system. I tried to compile the newer 3c90x, but couldn't get it to compile. However, this is likely not a problem, since the Ethernet-HOWTO says that my card (3c905B) is supported by this driver (3c59x). I don't know how to check if the card itself works, but the following messages (from dmesg) make me think it's ok: dmesg ... 3c59x.c:LK1.1.13 27 Jan 2001 Donald Becker and others. http://www.scyld.com/network/vortex.html See Documentation/networking/vortex.txt eth0: 3Com PCI 3c905B Cyclone 10/100/BNC at 0x1400, 00:50:04:0a:fb:13, IRQ 11 product code 5545 rev 00.0 date 01-21-99 8K byte-wide RAM 5:3 Rx:Tx split, autoselect/Autonegotiate interface. MII transceiver found at address 24, status 782d. Enabling bus-master transmits and whole-frame receives. eth0: scatter/gather enabled. h/w checksums enabled eth0: using NWAY device table, not 8
