Red Hat Bugzilla – Bug 58677
Incorrect network driver selected??
Last modified: 2014-08-11 01:46:01 EDT
Could not figure out which component should be used. So choose anaconda. It
does not like me putting anything in the Component text field.
Description of Problem:
Login to a newly installed system. Enable telnet. Telnet or ping another
node. Both fail. You get a no route to host error.
Version-Release number of selected component (if applicable):
Red Hat Linux release 7.1.90 (Roswell)
Kernel 2.4.9-14a on an alpha
How Reproducible: Very
This was tested on a DS 10
Steps to Reproduce:
1. Complete the installation.
2. Boot the machine and login.
3. Enable telnet. Telnet to or ping another system.
Telnet or ping to another node fails.
You can telnet to your self or ping your own node.
Successful login to another node.
After checking a little, did an lsmod and found that tulip driver was loaded.
Unloaded the tulip driver, loaded the de4x5 driver and restarted the network.
Now I could ping and telnet to any system.
What is the correct ethernet driver for the DS10 system?
Moved to version: beta 1 instead of RC1.
DS10 should be a de4x5
reason being that the foundry supplied chips are not 100% tulip correct.
Basically it get the MII interlink negotiation with the switch/hub wrong when
you use the tulip driver. de4x5 gets around this by selecting a generic
'catchall' interlink which is slower. The tulip driver just gets it wrong..