If you use cloonix, tell me if you feel it has got good perfs. From here I get 700KiloOctets/seconde to download the kernel from kernel.org into an uml machine in cloonix. What do you call "not hot perfs" ?
Of course the rest of the cloonix network is much better for local transmission with scp between host and uml virtual it is more than 4 mega to get into/outside an uml machine. As I use COW in my uml, when downloading big files into the machine, at the beginning the rate is good, but when the big file has to be really written in the COW, the bottleneck is the file writing, not the bandwidth. Le samedi 25 décembre 2010 à 11:58 -0800, Wil Barath a écrit : > Thanks for this, and I will use this scenario if I can't get slip > working. > > > Slirp's performance is not so hot, which is the only reason I'm > avoiding it. > > > Subject: Re: [uml-user] help getting slip networking going > > From: clow...@clownix.net > > To: w_bar...@hotmail.com > > CC: user-mode-linux-user@lists.sourceforge.net > > Date: Fri, 24 Dec 2010 23:59:08 +0100 > > > > Hello, > > I have no direct answer for the slip (I thought it war slirp) > > but the cloonix-8.01 embeds an uml openwrt backfire and an uml lenny > > and has its own slirp to go out to the internet without going > through > > the host'ip stack. > > see at cloonix.net or clownix.net for the set of tools to get a > running > > uml machine that is ready to surf the web. > > I hope this helps. > > > > > > > > Le vendredi 24 décembre 2010 à 12:09 -0800, Wil Barath a écrit : > > > I 've looked through the docs and googled the heck out of this > list > > > and the 'net, but can't find a single working example of a slip > > > setup. > > > > > > > > > I need to use slip because I am running the UML in a VPS (with no > > > kernel tun/tap support, exactly the situation pointed out in the > docs > > > as the only reason to use slip) > > > > > > > > > There's a lot of commentary about how slip was the first way UML > was > > > connected to the 'net and how it has drawbacks, but unfortunately > the > > > examples present in the documentation make a lot of assumptions > about > > > the host. I get the feeling the author genuinely believed that > using > > > slip was a bad thing, and so avoided giving a useful example as a > way > > > of discouraging its use. > > > > > > > > > > > > > > > Here's my config so far (all taken from the docs, but not > verbatim, as > > > it would collide with my upstream router) > > > > > > > > > #HOST ./linux-2.6.24-rc7 ubda=Fedora.cow,FedoraCore5-x86-root_fs > > > eth0=slip,192.168.0.1 > > > > > > > > > [... loading noise...] > > > > > > > > > [r...@localhost ~]# dmesg > > > Linux version 2.6.24-rc7-dirty (jd...@tp.user-mode-linux.org) (gcc > > > version 4.1.2 20070925 (Red Hat 4.1.2-27)) #97 Mon Jan 7 11:18:24 > EST > > > 2008 > > > [...snip...] > > > net_namespace: 64 bytes > > > Using 2.6 host AIO > > > NET: Registered protocol family 16 > > > NET: Registered protocol family 2 > > > Time: itimer clocksource has been installed. > > > Switched to high resolution mode on CPU 0 > > > IP route cache hash table entries: 1024 (order: 0, 4096 bytes) > > > TCP established hash table entries: 2048 (order: 2, 16384 bytes) > > > TCP bind hash table entries: 2048 (order: 1, 8192 bytes) > > > TCP: Hash tables configured (established 2048 bind 2048) > > > TCP reno registered > > > Checking host MADV_REMOVE support...<3>MADV_REMOVE failed, err = > -38 > > > Can't release memory to the host - memory hotplug won't be > supported > > > mconsole (version 2) initialized on /root/.uml/wOqluE/mconsole > > > [...snip...] > > > TCP cubic registered > > > NET: Registered protocol family 1 > > > NET: Registered protocol family 17 > > > Initialized stdio console driver > > > Console initialized on /dev/tty0 > > > console [tty0] enabled > > > Initializing software serial port version 1 > > > Choosing a random ethernet address for device eth0 > > > Netdevice 0 (66:35:f6:cf:fb:ab) : SLIP backend - SLIP IP = > 192.168.0.1 > > > console [mc-1] enabled > > > [...snip...] > > > Setting slip line discipline: Invalid argument > > > uml_net : waitpid process 15689 failed, errno = 10 > > > slip_tramp failed - err = 10 > > > > > > > > > [r...@localhost ~]# ifconfig > > > lo Link encap:Local Loopback > > > inet addr:127.0.0.1 Mask:255.0.0.0 > > > UP LOOPBACK RUNNING MTU:16436 Metric:1 > > > RX packets:0 errors:0 dropped:0 overruns:0 frame:0 > > > TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 > > > collisions:0 txqueuelen:0 > > > RX bytes:0 (0.0 b) TX bytes:0 (0.0 b) > > > > > > > > > [r...@localhost ~]# ifconfig eth0 up > > > Setting slip line discipline: Invalid argument > > > uml_net : waitpid process 27924 failed, errno = 10 > > > get_ifname failed, err = 22 > > > SIOCSIFFLAGS: Invalid argument > > > [r...@localhost ~]# route > > > Kernel IP routing table > > > Destination Gateway Genmask Flags Metric Ref > > > Use Iface > > > 169.254.0.0 * 255.255.0.0 U 0 0 > > > 0 lo > > > > > > > > > [from the docs, 2 steps to configure the interface for slip, > albeit > > > not explicitly given.] > > > > > > > > > [r...@localhost ~]# cat /etc/sysconfig/network-scripts/ifcfg-eth0 > > > DEVICE=eth0 > > > ONBOOT=yes > > > IPADDR=192.168.0.100 > > > GATEWAY=192.168.0.1 > > > NETMASK=255.255.255.0 > > > > > > > > > [r...@localhost ~]# halt > > > > > > > > > #HOST ifconfig sl0 192.168.0.1 pointopoint 192.168.0.100 up > > > SIOCSIFADDR: No such device > > > sl0: unknown interface: No such device > > > SIOCSIFDSTADDR: No such device > > > sl0: unknown interface: No such device > > > sl0: unknown interface: No such device > > > > > > > > > Best guess at this point is I'm missing something here to create > the > > > virtual slip network interface, and possibly the other portion is > > > failing because it is not present, but the errors are not so > helpful. > > > "Setting slip line discipline: Invalid argument" "uml_net : > waitpid > > > process 27924 failed, errno = 10" "get_ifname failed, err = 22" > > > "SIOCSIFFLAGS: Invalid argument" - sounds like the guest side (or > > > echoing the host side, vague) was unable to find the slip > interface. > > > > > > > > > > > > > > > > ------------------------------------------------------------------------------ > > > Learn how Oracle Real Application Clusters (RAC) One Node allows > customers > > > to consolidate database storage, standardize their database > environment, and, > > > should the need arise, upgrade to a full multi-node Oracle RAC > database > > > without downtime or disruption > > > http://p.sf.net/sfu/oracle-sfdevnl > > > _______________________________________________ > User-mode-linux-user mailing list > User-mode-linux-user@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user > > > > > ------------------------------------------------------------------------------ Learn how Oracle Real Application Clusters (RAC) One Node allows customers to consolidate database storage, standardize their database environment, and, should the need arise, upgrade to a full multi-node Oracle RAC database without downtime or disruption http://p.sf.net/sfu/oracle-sfdevnl _______________________________________________ User-mode-linux-user mailing list User-mode-linux-user@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/user-mode-linux-user
