Bernard Li wrote:
> Hi Andrea:
>
> I rebuilt trunk with your updates and monitoring seems to work again,
> however I am still having network issues. I have an Intel PRO/100+
> NIC on the client node and for whatever reason, the following is the
> output of dmesg | grep Intel:
>
> Intel(R) PRO/1000 Network Driver - version 6.3.9-k4-NAPI Copyright
> (c) 1999-2005 Intel Corporation. Intel(R) PRO/10GbE Network Driver -
> version 1.0.100-k2 Copyright (c) 1999-2005 Intel Corporation. e100:
> Intel(R) PRO/100 Network Driver, 3.5.10-k2-NAPI e100: Copyright(c)
> 1999-2005 Intel Corporation Intel ISA PCIC probe: not found.
>
> Is that normal? It appears that multiple drivers were loaded. Also,
> I think this is not a module as I don't see it in "lsmod" - so this
> is probably loaded in kernel?
That's because all the available networks drivers are compiled
statically into the kernel.. but the kernel should be able to probe and
use only the correct one...
As discussed with Brian the network drivers are needed into the kernel
to reach the image server, then additional supports are provided by the
boel_binaries.tar.gz.
Another solution could be to compile all as modules and store all that
modules into the initrd, but the big disadvantage of this approach
reduces the scalability. In fact kernel+initrd can be pulled by the
clients only via TFTP, so they can't exploit other nice transports, like
multicast or bittorrent.
There would be a 3rd solution... compile all as modules and split the
network modules (stored into the initrd) and the other ones (stored in
boel_binaries)... it should not be so difficult but it needs some work
to do... what do you think?
>
> BTW, I'd like to suggest the following patch for si_monitortk - minor
> text formatting that makes it look a bit more condensed:
It seems ok... go and check it. ;-)
>
> Index: sbin/si_monitortk
> ===================================================================
> --- sbin/si_monitortk (revision 3480) +++ sbin/si_monitortk
> (working copy) @@ -681,10 +681,10 @@ } else { if
> ($client->{$mac}->{'speed'} < 1000) { $client->{$mac}->{'speed'} = -
> sprintf("%.3fKB/s", $client->{$mac}->{'speed'}); +
> sprintf("%.0fKB/s", $client->{$mac}->{'speed'}); } else {
> $client->{$mac}->{'speed'} = - sprintf("%.3fMB/s",
> $client->{$mac}->{'speed'} / 1000); +
> sprintf("%.2fMB/s", $client->{$mac}->{'speed'} / 1000); } } # Format
> the total amount of RAM.
>
> I think the function "autodetect_hardware_and_load_modules()" in
> "functions" is no longer necessary, I'll work on removing that as
> well as other references to "discover".
>
This feature now should be included into the hotplug stuff... but I'm
not sure... we need some tests...
Cheers,
-Andrea
-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting language
that extends applications into web and mobile media. Attend the live webcast
and join the prime developer group breaking into this new coding territory!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
_______________________________________________
Sisuite-devel mailing list
Sisuite-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sisuite-devel
