LinuxQuestions.org - Bringing up interface eth0 determining IP information fails

- Linux - Newbie (https://www.linuxquestions.org/questions/linux-newbie-8/)

- - Bringing up interface eth0 determining IP information fails (https://www.linuxquestions.org/questions/linux-newbie-8/bringing-up-interface-eth0-determining-ip-information-fails-838561/)

Bringing up interface eth0 determining IP information fails

CentOS 4.7 a minimal installation with no GUI

Ethernet Controller RTL-8029

Running /sbin/lspci shows the Ethernet card....

ifcfg-eth0 shows BOOTPROTO=dhcp

NIC is connected to a LINKSYS WRT54G running DHCP. There are plenty of available IP assignments. All other PC that I have connected to the LINKSYS work fine. The CAT 5 cable is fine.

Why is this NIC not taking a DHCP assignment ?

what is the output of the following?

dhclient eth0

and then

ifconfig eth0

DHCPDISCOVER on NUMEROUS ports, then No DHCPOFFERS received. Np working leases in persistent database - sleeping

ifconfig eth0 shows the interface but no inet IP assignment. Also to note is 113 dropped RX packets and 8 errors on TX packets

Quote:

Originally Posted by johnshen64 (Post 4129942)

what is the output of the following?

dhclient eth0

and then

ifconfig eth0

Hi -

From the sounds of it, it *should* work. A couple of observations/suggestions:

1. It sounds like you're knowledgeable, and that you're doing everything right.

2. You imply you've got an Ethernet adapter (not wifi), and you're plugging it directly into your WRT54G router.

Q: is this correct?

3. If so, please double-check your /sbin/ifconfig output, and DISABLE all interfaces besides your eth0. Don't mess with multiple interfaces until we get *one* interface working.

4. The problem clearly appears to be DHCP.

Q: Have you considered simply assigning a static address (if only for debugging purposes)?

I have quite a few years of technical experience so Im not exactly a newbie ;-)

Yes, I have a Ethernet card with a CAT 5 cable connected to the router.

eth0 is the only interface at this point aside from local loopback.

Static assignment ? I still want to understand why assignment is not happening dynamically. If need be, we can go that route.

Quote:

Originally Posted by paulsm4 (Post 4129978)

Well, you can always try a network trace to see who's dropping the ball. You can use tcpdump (cmd-line) on Centos, or Wireshark on any of the other PCs on your LAN.

There might be some kind of conflict between Netmanager and your low-level configuration files (such as /etc/sysconfig/network-scripts/ifcfg-eth0, which I believe you cited in your original post). Who knows - Centos might also have some firewall setting that's blocking DHCP. But I think I'm stretching...

If it's no hassle, try seeing if a static address works.

And if so, double-check the settings in your /etc/sysconfig/network-scripts and/or go back to DHCP and try taking a network trace.

Good luck - and please post back what you find!

PS:
1. FIREWALL EXAMPLE:

Code:

iptables -I INPUT -i eth0 -p udp --sport 67:68 --dport -j ACCEPT

  <= Insure DHCP isn't getting blocked

2. NETCONF EXAMPLE:
vi /etc/sysconfig/network-scripts/ifcfg-eth0

Code:

# Example static: DON'T do this!

# DEVICE=eth0

# BOOTPROTO=static

# HWADDR=00:19:D1:2A:BA:A8

# IPADDR=10.10.29.66

# NETMASK=255.255.255.192

# ONBOOT=yes



# Example DHCP: DON'T specify IP or subnet mask...

DEVICE=eth0

BOOTPROTO=dhcp

HWADDR=00:19:D1:2A:BA:A8

ONBOOT=yes

Hi, Stephanie -

Q: Did you get it working?

Q: If so, what was the problem?
If not, perhaps maybe I could install CentOS 4.7 on a VM and try reproducing it? (I'm currently running CentOS 5.5 on a VPS, but that's not really applicable to your scenario).

Please let us know!

TIA .. PSM

The card may nor may not be configured for auto negotiation. Does mii-tool or ethtool recognize the adapter? If so how is it configured?

Nope, I upgraded to CentOS 4.8 because my application allowed that. I acutally ordered a new ethernet card that is stated to be compatible with Linux. My current card uses a RTL8029 chipset which is a commonly used chipset though.

Quote:

Originally Posted by paulsm4 (Post 4131816)

The adapter appears to be recognized as one can configure it manually. Tell me more about these tools! I am new to these

Quote:

Originally Posted by michaelk (Post 4131837)

The card may nor may not be configured for auto negotiation. Does mii-tool or ethtool recognize the adapter? If so how is it configured?

OK, then please cut/paste the exact output of the following:

1. ifconfig eth0

Code:

  /sbin/ifconfig eth0

eth0      Link encap:Ethernet  HWaddr 00:0C:76:E2:71:C5

          inet addr:192.168.1.xxx  Bcast:192.168.1.xxx  Mask:255.255.255.0

          inet6 addr: fe80::xxx:76ff:fee2:71c5/64 Scope:Link

          UP BROADCAST NOTRAILERS RUNNING MULTICAST  MTU:1500  Metric:1

          RX packets:593541 errors:0 dropped:0 overruns:0 frame:0

          TX packets:386699 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:1000

          RX bytes:88323609 (84.2 Mb)  TX bytes:487001350 (464.4 Mb)

          Interrupt:10 Base address:0xcf00

2. dhclient eth0

Code:

/sbin/dhclient eth0

Internet Systems Consortium DHCP Client V3.0.5-RedHat

Copyright 2004-2006 Internet Systems Consortium.

All rights reserved.

For info, please visit http://www.isc.org/sw/dhcp/



/sbin/dhclient-script: configuration for eth0 not found. Continuing with defaults.

/etc/sysconfig/network-scripts/network-functions: line 78: eth0: No such file or directory

SIOCSIFADDR: No such device

eth0: unknown interface: No such device

eth0: unknown interface: No such device

Failed to get interface index: No such device

3. ethtool eth0 ("ethtool" is generally preferred over "mii-tool"):

Code:

/usr/sbin/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

4. less /var/log/messages
<= Cut/paste any messages related to the failed DHCP request(s)

... and, most importantly ...

5. dhcpcd -r

Code:

/sbin/dhcpcd -r

****  /sbin/dhcpcd: already running

****  /sbin/dhcpcd: if not then delete /var/run//dhcpcd-eth0.pid file

Getting information off the machine onto this machine is a bit of an issue.

I can summarize the results this way:

ifconfig eth0: shows no assigned inet addr, states UP BROADCAST RUNNING MULTICAST, 18 RX packet errors & 7 TX packet errors

dhclient eth0: shows several DHCPDISCOVER attempt messages and finally NO DHCP OFFERS received

ethtool eth0: showed results similar to your "code" results. Actually, the RTL8029 NIC showed "NO DATA AVAILABLE" after running ethtool. My new DLINK NIC showed a whole table of results.

less /var/log/messages: showed ifup failed. Bringing up interfae eth0 failed. Saw a number of NETDEV WATCHDOG eth0 transmit timed out messages

dhcpcd does not exist in sbin...this is the client daemon ? If I type dhclient again it tells me its already running.

Quote:

Originally Posted by paulsm4 (Post 4132553)

OK, then please cut/paste the exact output of the following:

1. ifconfig eth0

Code:

  /sbin/ifconfig eth0

eth0      Link encap:Ethernet  HWaddr 00:0C:76:E2:71:C5

          inet addr:192.168.1.xxx  Bcast:192.168.1.xxx  Mask:255.255.255.0

          inet6 addr: fe80::xxx:76ff:fee2:71c5/64 Scope:Link

          UP BROADCAST NOTRAILERS RUNNING MULTICAST  MTU:1500  Metric:1

          RX packets:593541 errors:0 dropped:0 overruns:0 frame:0

          TX packets:386699 errors:0 dropped:0 overruns:0 carrier:0

          collisions:0 txqueuelen:1000

          RX bytes:88323609 (84.2 Mb)  TX bytes:487001350 (464.4 Mb)

          Interrupt:10 Base address:0xcf00

2. dhclient eth0

Code:

/sbin/dhclient eth0

Internet Systems Consortium DHCP Client V3.0.5-RedHat

Copyright 2004-2006 Internet Systems Consortium.

All rights reserved.

For info, please visit http://www.isc.org/sw/dhcp/



/sbin/dhclient-script: configuration for eth0 not found. Continuing with defaults.

/etc/sysconfig/network-scripts/network-functions: line 78: eth0: No such file or directory

SIOCSIFADDR: No such device

eth0: unknown interface: No such device

eth0: unknown interface: No such device

Failed to get interface index: No such device

3. ethtool eth0 ("ethtool" is generally preferred over "mii-tool"):

Code:

/usr/sbin/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

4. less /var/log/messages
<= Cut/paste any messages related to the failed DHCP request(s)

... and, most importantly ...

5. dhcpcd -r

Code:

/sbin/dhcpcd -r

****  /sbin/dhcpcd: already running

****  /sbin/dhcpcd: if not then delete /var/run//dhcpcd-eth0.pid file

Sorry :-(

You're *so* doing all the right things - it's frustrating that it hasn't paid off yet.

If at all possible, please try a static IP. If only for debugging purposes.

IMHO...

PS:
Some "grasping at straws" suggestions:

1. Disable your Linux firewall, e.g.
service iptables stop

2. Disable selinux (frankly, if you're at Centos 4.x, I doubt you even *have* selinux):
setenforce 0

3. You could do worse than run a tcpdump trace on Centos, and/or a Wireshark trace anywhere else

It is frustrating...wah ! lol

I tried setting a static IP. Set a default GW. Flushed iptables of its rules. Still cant ping beyond my NIC.

Have any good primers on interpreting tcpdump ?

Quote:

Originally Posted by paulsm4 (Post 4133127)

Hi -

1. At its simplest level (which is absolutely all you need), "tcpdump" is really, really easy:

Quote:

tcpdump -nS

Here's an example:

Code:

tcpdump -nS > tmp.txt

<<wait a few moments, then Ctl-C>>



less tmp.txt

10:55:10.193469 IP zz.183.40.xxx.49985 > zz.99.83.xxx.ssh: . ack 3499409730 win 32920

10:55:10.193509 IP zz.99.83.xxx.ssh > zz.183.40.xxx.49985: P  499409846:3499410146(300) ack 4003329759 win 8576

10:55:10.242646 IP zz.183.40.xxx.49985 > zz.99.83.xxx.ssh: . ack  499410146 win 32504

10:55:12.499468 IP zz.183.40.xxx.49985 > zz.99.83.xxx.ssh: P  003329759:4003329811(52) ack 3499410146 win 32504

2. Here are two tcpdump links:
http://www.tcpdump.org/
http://danielmiessler.com/study/tcpdump/

3. Any luck on the "grasping at straws" suggestions (firewall and/or SELinux)?

4. Any chance to try swapping hardware (swap cable, swap NIC, try on a side-by-side PC, etc)?

Crossing my fingers :)