[K12OSN] K12LTSP 5.0.0 Release Candidate #4

[Edward Holcroft]

>>>
>>> At the BGC of Hawaii, I just updated our FC3-based install.  Yes,  
>>> after more than two years of heavy teenage use, this lab is still  
>>> chugging.  I did a clean install of K12LTSP5rc1, and all but two  
>>> of the clients rebooted with the following message
>>>
>>> RAMDISK: Couldn't find valid RAM disk image starting at 0.
>>> Kernel panic - not syncing: VFS: Unable to mount root fs on  
>>> unknown-block(1,0)
>>>
>>> What's perplexing is that all the clients are in the same switch  
>>> and are the same architecture.  I am updating the packages  
>>> tonight, but as it stands the default install has some issues.
>>>
>>
>> I am getting this same error and will be very interested to know  
>> how you solve it. Will run a yum update today to see if that  
>> resolves anything. I found that it occurs on clients with rtl8139  
>> etherboot cards. My 3c905 PXE cards work fine.
>>
>
> Just did a yum update to RC4 but I am still getting the error. Same  
> error with different types of clients from P1's through to P3's.  
> The only consistent component is the rtl8139 NIC, which seems  
> consistently to produce the kernel panic. Tried a whole bunch of  
> cards too, so it's not just one bad card producing an error. These  
> same clients all boot just fine off a 4.4.1 server.
>
Apologies for replying to myself, but I have just tested a thin  
client with Etherboot running on a natsemi 83815 and RC4. It  
idisplays the same issue. I have an Intel Pro100 and a Via Rhine that  
work happily on RC4 using the same client boxes, so I can rule out  
the clients as the root of the problem, as well as a specific rtl8139  
issue.

I am using 5.0.7 and 5.0.10 of Etherboot on these NICs. Is there a  
chance that the older Etherboot is incompatible with the 2.6 LTSP  
kernel?

cheers
ed