From: TJ@noneofyour.business   
      
   On 09/27/2012 12:53 AM, Adam wrote:   
   > [I'm replying to all three posts dealing with network setup (and nothing   
   > else) as they overlap. This also creates a "new" subthread dealing only   
   > with that.]   
   >   
   > Moe Trin wrote:   
   >> On Mon, 24 Sep 2012, in the Usenet newsgroup alt.os.linux.mandriva, in   
   >> article   
   >> , Adam wrote:   
   >>   
   >>> I'd gotten Wi-Fi working for two distros (Fedora 17 and CentOS 6.3)   
   >>> so I'm using one of those to send this, but haven't yet gotten   
   >>> either a wired or wireless connection on it with Mageia 2 (or any of   
   >>> the other distros on it).   
   >   
   > "Other distros" meaning, right now, Mageia 2 and Debian 6.0.5 which are   
   > running on eris but don't have any 'net connection, although I'm trying   
   > (hoping?) for a wired connection ATM. Interface of choice is currently   
   > wlan0 on the two with working wireless but eth0 on the two with no   
   > connection yet.   
   >   
   >> Getting the wireless running sounds like an uncooperative "helper"   
   >> program.   
   >   
   > TJ said the updates to Mageia 2 include a better "wireless config"   
   > helper. Of course I can't get the updates until it has some kind of   
   > working connection. Right now I'm working on getting a wired connection   
   > solely to get those updates.   
   >   
   Um, I merely said the tools had been updated. That should mean that   
   they're "better," but being even more ignorant on the subject than I   
   thought I was, I can't say for sure.   
      
   >> See the boot messages to see what the working distros are   
   >> using as the driver module for the wireless.   
   >   
   > All distros are using rt2661.bin from the Ralink website, so that's   
   > probably not the problem.   
   >   
   >> As for the wired connection, does the router   
   >> know to hand out an IP to the hardware address of "eris"?   
   >   
   > My router now knows about both "eris" (192.168.1.23), the wireless   
   > connection, and "eris-wired" (192.168.1.24), the wired connection. I   
   > decided to do it that way because in about two weeks, eris will be going   
   > back to the other room where it only has a wireless connection. For   
   > /etc/sysconfig/network, I've created both -wired and -wireless versions,   
   > and /etc/sysconfig/network is a symlink, to network-wired right now.   
   >   
   >> Does /sbin/ifconfig show an eth0 interface with   
   >> reasonable IP address on _ALL_ distributions?   
   >   
   > Fedora doesn't have eth0 because wlan0 works properly, but it does have   
   > an interface called "p32p1". CentOS has eth0 with no IP address because   
   > its wlan0 also works. Debian and Mageia have eth0 correctly showing   
   > 192.168.1.24 and wlan0 with no IP address. Debian also has an interface   
   > called "pan0".   
   >   
   >> Does /sbin/route show reasonable data on _ALL_ distributions?   
   >   
   > Yes. Mageia shows:   
   >   
   > Destination Gateway Genmask Flags Metric Ref Use Iface   
   > 0.0.0.0 192.168.1.13 0.0.0.0 UG 5 0 0 eth0   
   > 169.254.0.0 0.0.0.0 255.255.0.0 U 5 0 0 eth0   
   > 192.168.1.0 0.0.0.0 255.255.255.192 U 5 0 0 eth0   
   >   
   > where 169.254.0.0 is named "link-local". The other three have:   
   >   
   > Destination Gateway Genmask Flags Metric Ref Use Iface   
   > 192.168.1.0 0.0.0.0 255.255.255.192 U 0 0 0 eth0   
   > 0.0.0.0 192.168.1.13 0.0.0.0 UG 0 0 0 eth0   
   >   
   > with the lines sometimes in the opposite order, and the two distros with   
   > working wireless have Iface "wlan0" instead of "eth0". Sounds   
   > reasonable to me.   
   >   
   >> Can you use /sbin/ifconfig   
   >> and /sbin/route to MANUALLY bring up the interface on those distros   
   >> that don't show the desired interface?   
   >   
   > Didn't try, because ifconfig on all distros shows the desired interface   
   > with the correct IP address.   
   >   
   >> What does /bin/hostname show   
   >> and how does that name relate to the network interfaces?   
   >   
   > IIRC eris.aplomb.invalid on the two with a working wireless connection   
   > and eris-wired.aplomb.invalid on the two where I'm trying to set up a   
   > wired connection.   
   >   
   >> possible confusion between the names in /etc/hosts   
   >> and the output of /bin/hostname.   
   >   
   > That's straightened out.   
   >   
   > [root@eris-wired ~]# cat /etc/hosts   
   > 127.0.0.1 localhost.localdomain localhost   
   > 192.168.1.13 dslrouter.aplomb.invalid dslrouter   
   > 192.168.1.23 eris.aplomb.invalid eris   
   > 192.168.1.24 eris-wired.aplomb.invalid eris-wired   
   > 192.168.1.33 HPCLJ5.aplomb.invalid HPCLJ5 hpclj5   
   > 192.168.1.43 stolid.aplomb.invalid stolid   
   > [root@eris-wired ~]#   
   >   
   >> Above, you say 192.168.1.23 for wireless, 192.168.1.24 for Ethernet.   
   >> How is the eth0 interface being configured? Static setup from files   
   >> on 'eris' (possibly /etc/sysconfig/network)? Static IP provided   
   >> from router? DHCP from router?   
   >   
   > All static IP from files for each distro on eris.   
   >   
   FWIW, my router, a Linksys WRT54G, is set to DHCP, so that an authorized   
   "guest" could connect more easily. Numbers that are handed out are   
   remembered for 24 hours after disconnection before being made available   
   again. I have the wired computer set to get the IP from the router, but   
   my two wireless connections are set up with static IPs within the   
   distro, within range but high enough that it's doubtful that such a   
   number would be handed out.   
      
      
   >>> RX packets:6 errors:0 dropped:0 overruns:0 frame:0   
   >>> TX packets:23 errors:0 dropped:0 overruns:0 carrier:0   
   >>   
   >> OK - link is up and working (packet counts, and no errors)   
   >   
   > Interface of choice on all four distros reports packets both sent and   
   > received without any errors. Sometimes other interfaces report a few   
   > packets both ways, again without errors.   
   >   
   > BTW I just thought to check /etc/resolv.conf and all four look about the   
   > same:   
   >   
   > # Generated by NetworkManager   
   > search aplomb.invalid   
   > nameserver 68.237.161.12   
   > nameserver 71.243.0.12   
   >   
   > I'm temporarily using Verizon's DNS, and they obviously work correctly.   
   >   
   >>> What may be relevant is that for a while earlier today, using Mageia's   
   >>> wired connection, attempts to ping local hosts were completely   
   >>> successful, while attempts to ping remote hosts (e.g. www.google.com)   
   >>> reported around 65-70% packet loss.   
   >>   
   >> Any error messages in applications?   
   >   
   > When I point a browser to http://192.168.1.13 (my router's main menu),   
   > eventually I get "took too long to respond" or a similar timeout error.   
   >   
   > What confuses me most of all is the 40-80% packet loss rate when pinging   
   > anything else, even my own router. What could be causing an   
   > intermittent problem? I'd expect either close to 0% or 100% packet loss.   
   >   
   >> Don't post it, but look at the   
      
   [continued in next message]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)   
|