Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
2013/8/12 Alan McKinnon alan.mckin...@gmail.com: On 12/08/2013 09:13, gevisz wrote: The response of the first router contained an error that prevented all the other applications to use it, the system knew about it (for example from the output of the host utility) but, nevertheless did not proceeded with the next router listed in resolv.conf. I do undersand that this may be because of the layered structure of the networked software. But, nevertheless, I think that something is fundamentally wrong with this. What kind of error did you get? As I have already wrote it earlier, with three different DNS in /etc/resolv.conf and /etc/conf.d/net files, the host utility correctly reported IP address of a site (eg, www.google.com) but added the following message: ;; Warning: query response not set With only the first (my local DNS) in /etc/resolv.conf and /etc/conf.d/net files, the output of the host utility was as follows: # host www.google.com www.google.com has address 74.125.232.52 www.google.com has address 74.125.232.48 www.google.com has address 74.125.232.49 www.google.com has address 74.125.232.50 www.google.com has address 74.125.232.51 ;; Warning: query response not set ;; Warning: query response not set Host www.google.com not found: 4(NOTIMP) In both cases above no internet application (eg, links or firefox) could convert site names to IP adresses and only after deleting the first (local) DNS from /etc/resolv.conf and /etc/conf.d/net files, internet applications started to work as expected (and the host utility, in this case, returned no error or warning message) That have proved to myself that The response of the first router contained an error that prevented all the other applications to use it, the system knew about it (for example from the output of the host utility) but, nevertheless, did not proceeded with the next router listed in resolv.conf [or /etc/conf.d/net]. I do undersand that this may be because of the layered structure of the networked software. But, nevertheless, I think that something is fundamentally wrong with this. If complete garbage came back, I'm not sure what the resolver does with that (oddly enough, I never tested that) The more usual case is you get a proper DNS result of NXDOMAIN which indicates the query is valid, but the entry is not in DNS. It's pointless trying another cache as per DNS, they should all then return that result. This is why the router did not try the other entries in resolv.conf - that usually only happens when a cache does not respond. So the behaviour you saw is probably correct albeit not the behaviour you wanted. -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On 13/08/2013 08:31, gevisz wrote: 2013/8/12 Alan McKinnon alan.mckin...@gmail.com: On 12/08/2013 09:13, gevisz wrote: The response of the first router contained an error that prevented all the other applications to use it, the system knew about it (for example from the output of the host utility) but, nevertheless did not proceeded with the next router listed in resolv.conf. I do undersand that this may be because of the layered structure of the networked software. But, nevertheless, I think that something is fundamentally wrong with this. What kind of error did you get? As I have already wrote it earlier, with three different DNS in /etc/resolv.conf and /etc/conf.d/net files, the host utility correctly reported IP address of a site (eg, www.google.com) but added the following message: ;; Warning: query response not set With only the first (my local DNS) in /etc/resolv.conf and /etc/conf.d/net files, the output of the host utility was as follows: # host www.google.com www.google.com has address 74.125.232.52 www.google.com has address 74.125.232.48 www.google.com has address 74.125.232.49 www.google.com has address 74.125.232.50 www.google.com has address 74.125.232.51 ;; Warning: query response not set ;; Warning: query response not set Host www.google.com not found: 4(NOTIMP) In both cases above no internet application (eg, links or firefox) could convert site names to IP adresses and only after deleting the first (local) DNS from /etc/resolv.conf and /etc/conf.d/net files, internet applications started to work as expected (and the host utility, in this case, returned no error or warning message) That have proved to myself that The response of the first router contained an error that prevented all the other applications to use it, the system knew about it (for example from the output of the host utility) but, nevertheless, did not proceeded with the next router listed in resolv.conf [or /etc/conf.d/net]. I do undersand that this may be because of the layered structure of the networked software. But, nevertheless, I think that something is fundamentally wrong with this. the host command is not your local resolver in libc, you cannot take the output of host and conclude anything about your resolver, as they operate in fundamentally different ways with entirely different purposes. Your DNS setup is doing exactly what it is supposed to do - the first cache returned an error and your local resolver concludes the query cannot be resolved, so stops trying. Solution: upgrade your router's firmware. It looks like it's bust. If complete garbage came back, I'm not sure what the resolver does with that (oddly enough, I never tested that) The more usual case is you get a proper DNS result of NXDOMAIN which indicates the query is valid, but the entry is not in DNS. It's pointless trying another cache as per DNS, they should all then return that result. This is why the router did not try the other entries in resolv.conf - that usually only happens when a cache does not respond. So the behaviour you saw is probably correct albeit not the behaviour you wanted. -- Alan McKinnon alan.mckin...@gmail.com -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
I somehow missed this post, so excuse me for the late reply. 2013/8/5 Mick michaelkintz...@gmail.com: On Monday 05 Aug 2013 07:06:08 gevisz wrote: My thanks to all who replied to my question. The problem was with my local router, which I also used as DNS. After excluding it from /etc/resolv.config and /etc/init.d/net files, Firefox started to work as expected. Hmm ... I wonder if this is related to my earlier comment about malformed packets. Somewhere, you hinted that the problem may be with the routers and suggested to experiment with it. Before that, I strongly believed that, if I listed 3 different routers in my resolv.conf, the system should proceed with the next router if something is wrong with the previous one, but unfortunately it did not. The response of the first router contained an error that prevented all the other applications to use it, the system knew about it (for example from the output of the host utility) but, nevertheless did not proceeded with the next router listed in resolv.conf. I do undersand that this may be because of the layered structure of the networked software. But, nevertheless, I think that something is fundamentally wrong with this. Once more, thank you for your help. A few following remarks are minor and so, you can stop your reading here. May be worth trying a different firmware for this router. I have already changed the firmware after purchasing it but now I cannot afford it as I need its uninterupted functioning. Suggestions of Michael Kintzios This is the new kernel naming scheme of NICs. Which-ever nomenclature you decide to use, check that that's the only one having a symlink in /etc/init.d to net.lo Yes, there is only enp2s15 links to lo in /etc/init.d The idea here is that you need consistent naming of your iface. If you have settled on the kernel naming of enp2s15, then stick with this throughout your configuration. Yes, I did. After deleting all but my lan router DNS from /etc/conf.d/net and /etc/resolv.conf files, I had the same problem as before but, in addition, the host utility reports an additional error. Please, see the full response below. You should not need to manually alter anything in your /etc/resolv.conf, which will be completed with the DNS server name(s) you have set up in your /etc/conf.d/net. Actually, I changed it in both files simultaneously, but -- as I have already explained it above, yes, I should not do it but had to. :^) # host www.google.com www.google.com has address 74.125.232.52 www.google.com has address 74.125.232.48 www.google.com has address 74.125.232.49 www.google.com has address 74.125.232.50 www.google.com has address 74.125.232.51 ;; Warning: query response not set ;; Warning: query response not set I think this means that the DNS server response is incorrectly formed (or that the server respond code does not include a 4 bit RCODE as it should - more detail for DNS geeks can be found here: http://www.ietf.org/rfc/rfc2136.txt) Thank you, for the referrence. I will study it later. Host www.google.com not found: 4(NOTIMP) The RFC says: The name server does not support the specified Opcode. I would reflash the firmware, or try any OpenSource alternatives if available for your router. It is a small router device. I have already changed its firmware after purchasing it to a newer one. I do not know if its open source alternative exists and, anyway, I cannot change it now because I cannot afford any interruption of the router functioning. After leaving in /etc/conf.d/net and /etc/resolv.conf files only the DNS of my service provider, Firefox started to work as predicted. Thank you! This may not be ideal (it will introduce some latency in your requests) but if you can't fix your router, it'll have to do for now. Can you please show us: ip route show ip addr show ip link show $ ip route show default via 192.168.0.1 dev enp2s15 metric 2 127.0.0.0/8 via 127.0.0.1 dev lo scope link 192.168.0.0/24 dev enp2s15 proto kernel scope link src 192.168.0.9 This says that your IP address us 192.168.0.9, but see below. $ ip addr show [snip ...] 2: enp2s15: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether MAC_address_of_my_Ethernet_card brd ff:ff:ff:ff:ff:ff inet 192.168.0.7/24 brd 192.168.0.255 scope global enp2s15 This says that your ip address is 192.168.0.7 - did you get a different IP address between the two commands? Your /etc/conf.d/net showed that you had set up a static address as config_enp2s15=192.168.0.9 ... so why is this here? Sorry, it happened only because of my stupid attempt to eliminate all the real IP addresses... $ ip link show [snip ...] 2: enp2s15: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc pfifo_fast state UP mode DEFAULT qlen 1000 link/ether MAC_address_of_my_Ethernet_card brd ff:ff:ff:ff:ff:ff OK, this looks good. Suggestions of Kurian
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On 12/08/2013 09:13, gevisz wrote: The response of the first router contained an error that prevented all the other applications to use it, the system knew about it (for example from the output of the host utility) but, nevertheless did not proceeded with the next router listed in resolv.conf. I do undersand that this may be because of the layered structure of the networked software. But, nevertheless, I think that something is fundamentally wrong with this. What kind of error did you get? If complete garbage came back, I'm not sure what the resolver does with that (oddly enough, I never tested that) The more usual case is you get a proper DNS result of NXDOMAIN which indicates the query is valid, but the entry is not in DNS. It's pointless trying another cache as per DNS, they should all then return that result. This is why the router did not try the other entries in resolv.conf - that usually only happens when a cache does not respond. So the behaviour you saw is probably correct albeit not the behaviour you wanted. -- Alan McKinnon alan.mckin...@gmail.com
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On 5 August 2013, at 18:28, Alan McKinnon wrote: ... So why change this? Because you can't rely on ethX always being the same physical hardware. On a firewall or router, you absolutely need to rely on this. The udev scheme works around this by letting you specify exact rules that will always do what you want. Why was this changed rammed down your throat? Well, that is political. The udev maintainers (along with systemd) work for Red Hat. RH's market is almost totally servers, and big multi-nic ones at that. They really need consistent names, doubly so if the host is a virtualization host. The catch: RH (or more exactly the udev maintainer employed by RH) probably couldn't give a toss what you think or want, and went ahead and fixed their problem expecting you to deal with it or shove off I believe this all stems from the rejection of BIOS dev names by Linus /or the kernel folks. This would have allowed the kernel to determine which interfaces were eth0 / eth1 from the BIOS / firmware of next gen server machines. An open standard to present this information through the firmware was agreed between at least Dell and one other major server vendor (HP springs to mind). The patches were rejected by the kernel folks because they risked renaming the interfaces on the small number of machines already in service with this firmware facility. I believe the response was do it in userspace, go talk to the udev guys, and the rest is history. Stroller.
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
My thanks to all who replied to my question. The problem was with my local router, which I also used as DNS. After excluding it from /etc/resolv.config and /etc/init.d/net files, Firefox started to work as expected. Thanks to all of you, I now have a working Gentoo on my desktop!!! Still, a small problem that vlc hangs the GNOME, but I hope to solve it in the nearest future (or post a separate question). Below is the full report of what I did reacting on your suggestions. Suggestions of Mark Pariente: Internet browsers (I have tried Firefox and Links) can connect to the WWW if I type the IP address as, for example, 173.194.71.104 but cannot if I type www.google.com And this is strange because the host utility works as expected (converting www.google.com into 173.194.71.104) just adding the following Warning: query response not set. Do other command line tools like ping work? Yes, the ping command works for everything I have tried including all the DNS I put into my resolv.conf file. It looks like your DNS setup is correct but for some reason the GNOME setup isn't. At first, I also blamed GNOME. However, after trying the Links web browser from a bare console (and obtaining the same results), it seems to me that the GNOME setup cannot be responsible for this problem Few things: Are you using NetworkManager? Is the 'networkmanager' USE flag enabled? If so try going to Settings-Network-Wired-gear icon-IPv4 and see what the DNS settings are like. I have emerged GNOME with the default set of USE flags and currently I have no NetworkManager in my GNOME menu. Also in Settings-Network, see if Network proxy is (mis)configured. My GNOME Network Proxy Preferences show the option Direct Internet Connection enabled. Suggestions of Michael Kintzios the host utility works as expected (converting www.google.com into 173.194.71.104) just adding the following Warning: query response not set. Hmm .. wonder if this is something to do with your dns repeater (your lan router) sending malformed packets in response? I am using Dlink DI-LB604 as my lan router and gateway. The ifconfig utility shows that instead of eth0 I have somewhat strange enp2s15 but I have played with this and found no difference after forcing udev to name it eth0. This is the new kernel naming scheme of NICs. Which-ever nomenclature you decide to use, check that that's the only one having a symlink in /etc/init.d to net.lo Yes, there is only enp2s15 links to lo in /etc/init.d try leaving only your router's local_network_IP in [your /etc/conf.d/net file] After deleting all but my lan router DNS from /etc/conf.d/net and /etc/resolv.conf files, I had the same problem as before but in addition the host utility reports an additional error. Please, see the full response below. # host www.google.com www.google.com has address 74.125.232.52 www.google.com has address 74.125.232.48 www.google.com has address 74.125.232.49 www.google.com has address 74.125.232.50 www.google.com has address 74.125.232.51 ;; Warning: query response not set ;; Warning: query response not set Host www.google.com not found: 4(NOTIMP) After leaving in /etc/conf.d/net and /etc/resolv.conf files only the DNS of my service provider, Firefox started to work as predicted. Thank you! Can you please show us: ip route show ip addr show ip link show $ ip route show default via 192.168.0.1 dev enp2s15 metric 2 127.0.0.0/8 via 127.0.0.1 dev lo scope link 192.168.0.0/24 dev enp2s15 proto kernel scope link src 192.168.0.9 $ ip addr show 1: lo: LOOPBACK,UP,LOWER_UP mtu 65536 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: enp2s15: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether MAC_address_of_my_Ethernet_card brd ff:ff:ff:ff:ff:ff inet 192.168.0.7/24 brd 192.168.0.255 scope global enp2s15 inet6 IPv6_address_of_my_Ethernet_card/64 scope link valid_lft forever preferred_lft forever 3: sit0: NOARP mtu 1480 qdisc noop state DOWN link/sit 0.0.0.0 brd 0.0.0.0 $ ip link show 1: lo: LOOPBACK,UP,LOWER_UP mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 2: enp2s15: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc pfifo_fast state UP mode DEFAULT qlen 1000 link/ether MAC_address_of_my_Ethernet_card brd ff:ff:ff:ff:ff:ff 3: sit0: NOARP mtu 1480 qdisc noop state DOWN mode DEFAULT link/sit 0.0.0.0 brd 0.0.0.0 Suggestions of Kurian Thayil Can you do a ping and see if the resolv.conf DNS ips are reachable? Yes, I can ping all my DNS. Moreover, I successfully use them from my Ubuntu installation on the same computer. do a dig @8.8.8.8 www.google.com ## which will do a name resolution with Google DNS servers. Here is the output: $ dig @8.8.8.8 www.google.co ; DiG
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On Monday 05 Aug 2013 07:06:08 gevisz wrote: My thanks to all who replied to my question. The problem was with my local router, which I also used as DNS. After excluding it from /etc/resolv.config and /etc/init.d/net files, Firefox started to work as expected. Hmm ... I wonder if this is related to my earlier comment about malformed packets. May be worth trying a different firmware for this router. Suggestions of Michael Kintzios This is the new kernel naming scheme of NICs. Which-ever nomenclature you decide to use, check that that's the only one having a symlink in /etc/init.d to net.lo Yes, there is only enp2s15 links to lo in /etc/init.d The idea here is that you need consistent naming of your iface. If you have settled on the kernel naming of enp2s15, then stick with this throughout your configuration. After deleting all but my lan router DNS from /etc/conf.d/net and /etc/resolv.conf files, I had the same problem as before but in addition the host utility reports an additional error. Please, see the full response below. You should not need to manually alter anything in your /etc/resolv.conf, which will be completed with the DNS server name(s) you have set up in your /etc/conf.d/net. # host www.google.com www.google.com has address 74.125.232.52 www.google.com has address 74.125.232.48 www.google.com has address 74.125.232.49 www.google.com has address 74.125.232.50 www.google.com has address 74.125.232.51 ;; Warning: query response not set ;; Warning: query response not set I think this means that the DNS server response is incorrectly formed (or that the server respond code does not include a 4 bit RCODE as it should - more detail for DNS geeks can be found here: http://www.ietf.org/rfc/rfc2136.txt) Host www.google.com not found: 4(NOTIMP) The RFC says: The name server does not support the specified Opcode. I would reflash the firmware, or try any OpenSource alternatives if available for your router. After leaving in /etc/conf.d/net and /etc/resolv.conf files only the DNS of my service provider, Firefox started to work as predicted. Thank you! This may not be ideal (it will introduce some latency in your requests) but if you can't fix your router, it'll have to do for now. Can you please show us: ip route show ip addr show ip link show $ ip route show default via 192.168.0.1 dev enp2s15 metric 2 127.0.0.0/8 via 127.0.0.1 dev lo scope link 192.168.0.0/24 dev enp2s15 proto kernel scope link src 192.168.0.9 This says that your IP address us 192.168.0.9, but see below. $ ip addr show [snip ...] 2: enp2s15: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether MAC_address_of_my_Ethernet_card brd ff:ff:ff:ff:ff:ff inet 192.168.0.7/24 brd 192.168.0.255 scope global enp2s15 This says that your ip address is 192.168.0.7 - did you get a different IP address between the two commands? Your /etc/conf.d/net showed that you had set up a static address as config_enp2s15=192.168.0.9 ... so why is this here? $ ip link show [snip ...] 2: enp2s15: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc pfifo_fast state UP mode DEFAULT qlen 1000 link/ether MAC_address_of_my_Ethernet_card brd ff:ff:ff:ff:ff:ff OK, this looks good. Suggestions of Kurian Thayil Can you do a ping and see if the resolv.conf DNS ips are reachable? Yes, I can ping all my DNS. Moreover, I successfully use them from my Ubuntu installation on the same computer. do a dig @8.8.8.8 www.google.com ## which will do a name resolution with Google DNS servers. Here is the output: $ dig @8.8.8.8 www.google.co ; DiG 9.9.2 @8.8.8.8 www.google.co ; (1 server found) ;; global options: +cmd ;; Got answer: ;; -HEADER- opcode: QUERY, status: NOERROR, id: 4036 ;; flags: qr rd ra; QUERY: 1, ANSWER: 12, AUTHORITY: 4, ADDITIONAL: 5 ;; OPT PSEUDOSECTION: ; EDNS: version: 0, flags:; udp: 4096 ;; QUESTION SECTION: ;www.google.co. IN A ;; ANSWER SECTION: www.google.co.86400 IN CNAME www3.l.google.com. www3.l.google.com.13 IN A 173.194.32.166 www3.l.google.com.13 IN A 173.194.32.167 www3.l.google.com.13 IN A 173.194.32.168 www3.l.google.com.13 IN A 173.194.32.169 www3.l.google.com.13 IN A 173.194.32.174 www3.l.google.com.13 IN A 173.194.32.160 www3.l.google.com.13 IN A 173.194.32.161 www3.l.google.com.13 IN A 173.194.32.162 www3.l.google.com.13 IN A 173.194.32.163 www3.l.google.com.13 IN A 173.194.32.164 www3.l.google.com.13 IN A 173.194.32.165 ;; AUTHORITY SECTION: google.com. 244594 IN NS ns3.google.com. google.com. 244594 IN NS ns2.google.com.
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On Mon, Aug 05, 2013 at 11:06:52AM +0100, Mick wrote: Suggestions of Michael Kintzios This is the new kernel naming scheme of NICs. Which-ever nomenclature you decide to use, check that that's the only one having a symlink in /etc/init.d to net.lo Yes, there is only enp2s15 links to lo in /etc/init.d The idea here is that you need consistent naming of your iface. If you have settled on the kernel naming of enp2s15, then stick with this throughout your configuration. -- Regards, Mick If this is the new kernel naming scheme of NICs, why this in dmesg: [4.725902] systemd-udevd[1176]: renamed network interface wlan0 to enp0s18f2u2 It looks as if systemd-udev renamed the NIC to me. Can you explain? Cheers, Bruce -- Happy Penguin Computers ') 126 Fenco Drive ( \ Tupelo, MS 38801 ^^ supp...@happypenguincomputers.com 662-269-2706 662-205-6424 http://happypenguincomputers.com/ A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail? Don't top-post: http://en.wikipedia.org/wiki/Top_post#Top-posting
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
Am Mon, 5 Aug 2013 07:59:09 -0500 schrieb Bruce Hill da...@happypenguincomputers.com: On Mon, Aug 05, 2013 at 11:06:52AM +0100, Mick wrote: Suggestions of Michael Kintzios This is the new kernel naming scheme of NICs. Which-ever nomenclature you decide to use, check that that's the only one having a symlink in /etc/init.d to net.lo Yes, there is only enp2s15 links to lo in /etc/init.d The idea here is that you need consistent naming of your iface. If you have settled on the kernel naming of enp2s15, then stick with this throughout your configuration. -- Regards, Mick If this is the new kernel naming scheme of NICs, why this in dmesg: [4.725902] systemd-udevd[1176]: renamed network interface wlan0 to enp0s18f2u2 It looks as if systemd-udev renamed the NIC to me. Can you explain? It already has been explained in the previous NIC renaming discussion: what's broken is renaming a device within the kernels internal namespace, which contains eth*, wlan* (and maybe others). The problem is that there is a race condition with the kernel when renaming ethX to ethY. What you *can* do is rename ethX to somethingelseX or somethingelseY, because then you are not racing against the kernel to hand out device names. This is explained on the website that also explains the new default renaming scheme used by udev. I (and IIRC others, too) already linked to it in in the old thread, and the relevant news item also referenced it, but here it is again: http://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/ HTH -- Marc Joliet -- People who think they know everything really annoy those of us who know we don't - Bjarne Stroustrup signature.asc Description: PGP signature
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On Mon, Aug 05, 2013 at 04:31:44PM +0200, Marc Joliet wrote: Am Mon, 5 Aug 2013 07:59:09 -0500 schrieb Bruce Hill da...@happypenguincomputers.com: If this is the new kernel naming scheme of NICs, why this in dmesg: [4.725902] systemd-udevd[1176]: renamed network interface wlan0 to enp0s18f2u2 It looks as if systemd-udev renamed the NIC to me. Can you explain? It already has been explained in the previous NIC renaming discussion: what's broken is renaming a device within the kernels internal namespace, which contains eth*, wlan* (and maybe others). The problem is that there is a race condition with the kernel when renaming ethX to ethY. What you *can* do is rename ethX to somethingelseX or somethingelseY, because then you are not racing against the kernel to hand out device names. This is explained on the website that also explains the new default renaming scheme used by udev. I (and IIRC others, too) already linked to it in in the old thread, and the relevant news item also referenced it, but here it is again: http://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/ The fact is that udev renamed the NIC. For the average Joe with one NIC (very large percentage of users) this is a non sequitur. For those of us with 2 or more NICs, myself included, we have already setup our systems to use multiple NICs for a purpose and configured the system so that nothing can/will/needs to rename subsequent NICs. My point is don't say the new kernel naming scheme of NICs, say the new systemd naming scheme of NICs. -- Happy Penguin Computers ') 126 Fenco Drive ( \ Tupelo, MS 38801 ^^ supp...@happypenguincomputers.com 662-269-2706 662-205-6424 http://happypenguincomputers.com/ A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail? Don't top-post: http://en.wikipedia.org/wiki/Top_post#Top-posting
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
Am Mon, 5 Aug 2013 09:41:09 -0500 schrieb Bruce Hill da...@happypenguincomputers.com: [...] My point is don't say the new kernel naming scheme of NICs, say the new systemd naming scheme of NICs. Oh, oops. I don't know how I managed to misread your email that spectacularly. Sorry! -- Marc Joliet -- People who think they know everything really annoy those of us who know we don't - Bjarne Stroustrup signature.asc Description: PGP signature
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On Monday 05 Aug 2013 15:41:09 Bruce Hill wrote: On Mon, Aug 05, 2013 at 04:31:44PM +0200, Marc Joliet wrote: Am Mon, 5 Aug 2013 07:59:09 -0500 schrieb Bruce Hill da...@happypenguincomputers.com: If this is the new kernel naming scheme of NICs, why this in dmesg: [4.725902] systemd-udevd[1176]: renamed network interface wlan0 to enp0s18f2u2 It looks as if systemd-udev renamed the NIC to me. Can you explain? It already has been explained in the previous NIC renaming discussion: what's broken is renaming a device within the kernels internal namespace, which contains eth*, wlan* (and maybe others). The problem is that there is a race condition with the kernel when renaming ethX to ethY. What you *can* do is rename ethX to somethingelseX or somethingelseY, because then you are not racing against the kernel to hand out device names. This is explained on the website that also explains the new default renaming scheme used by udev. I (and IIRC others, too) already linked to it in in the old thread, and the relevant news item also referenced it, but here it is again: http://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInte rfaceNames/ The fact is that udev renamed the NIC. For the average Joe with one NIC (very large percentage of users) this is a non sequitur. For those of us with 2 or more NICs, myself included, we have already setup our systems to use multiple NICs for a purpose and configured the system so that nothing can/will/needs to rename subsequent NICs. My point is don't say the new kernel naming scheme of NICs, say the new systemd naming scheme of NICs. Indeed! Thanks for bringing this to my attention. Here's my eth0: [6.437527] systemd-udevd[1407]: starting version 204 [7.457924] systemd-udevd[1428]: renamed network interface eth0 to enp11s0 while my wireless NIC stays named as always was (wlan0): [7.822350] b43-phy0: Broadcom 4312 WLAN found (core revision 15) [7.838741] b43-phy0: Found PHY: Analog 6, Type 5 (LP), Revision 1 [7.838760] b43-phy0 debug: Found Radio: Manuf 0x17F, Version 0x2062, Revision 2 [ 15.771370] b43-phy0: Loading firmware version 666.2 (2011-02-23 01:15:07) [ 15.775217] b43-phy0 debug: b2062: Using crystal tab entry 19200 kHz. [ 17.157109] b43-phy0 debug: Chip initialized [ 17.157427] b43-phy0 debug: 64-bit DMA initialized [ 17.157888] b43-phy0 debug: QoS disabled [ 17.167424] b43-phy0 debug: Wireless interface started [ 17.172410] b43-phy0 debug: Adding Interface type 2 [ 17.173097] IPv6: ADDRCONF(NETDEV_UP): wlan0: link is not ready BTW, I have no systemd installed, only udev-204 and udev-init-scripts-26. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On 5 August 2013, at 15:41, Bruce Hill wrote: … For those of us with 2 or more NICs, myself included, we have already setup our systems to use multiple NICs for a purpose and configured the system so that nothing can/will/needs to rename subsequent NICs. Except you can't say that if you have a bank of a thousand servers all net-booting a shared system image. Sys admins on the Dell Poweredge list were very pleased when they learned of the persistent device names plan. Stroller.
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On 05/08/2013 16:41, Bruce Hill wrote: On Mon, Aug 05, 2013 at 04:31:44PM +0200, Marc Joliet wrote: Am Mon, 5 Aug 2013 07:59:09 -0500 schrieb Bruce Hill da...@happypenguincomputers.com: If this is the new kernel naming scheme of NICs, why this in dmesg: [4.725902] systemd-udevd[1176]: renamed network interface wlan0 to enp0s18f2u2 It looks as if systemd-udev renamed the NIC to me. Can you explain? It already has been explained in the previous NIC renaming discussion: what's broken is renaming a device within the kernels internal namespace, which contains eth*, wlan* (and maybe others). The problem is that there is a race condition with the kernel when renaming ethX to ethY. What you *can* do is rename ethX to somethingelseX or somethingelseY, because then you are not racing against the kernel to hand out device names. This is explained on the website that also explains the new default renaming scheme used by udev. I (and IIRC others, too) already linked to it in in the old thread, and the relevant news item also referenced it, but here it is again: http://www.freedesktop.org/wiki/Software/systemd/PredictableNetworkInterfaceNames/ The fact is that udev renamed the NIC. For the average Joe with one NIC (very large percentage of users) this is a non sequitur. For those of us with 2 or more NICs, myself included, we have already setup our systems to use multiple NICs for a purpose and configured the system so that nothing can/will/needs to rename subsequent NICs. My point is don't say the new kernel naming scheme of NICs, say the new systemd naming scheme of NICs. Let me see if I can clarify somewhat. eth0 can be considered the kernel name - the kernel named the NIC according to it's own rules using the info it had available. Kernel names depend on discovery order and to a lesser degree on the kernel code (a dev could change how things are done for example) enp0s18f2u2 can be considered the userspace name - it's derived from the slot the card is plugged into, and is set by udev. The kernel doesn't really care about this stuff, but you might. Most people think of their NICS on multi-NIC machines in terms of positions i.e. third one on the left and can't work with whatever eth2 happens to be today So why change this? Because you can't rely on ethX always being the same physical hardware. On a firewall or router, you absolutely need to rely on this. The udev scheme works around this by letting you specify exact rules that will always do what you want. Why was this changed rammed down your throat? Well, that is political. The udev maintainers (along with systemd) work for Red Hat. RH's market is almost totally servers, and big multi-nic ones at that. They really need consistent names, doubly so if the host is a virtualization host. The catch: RH (or more exactly the udev maintainer employed by RH) probably couldn't give a toss what you think or want, and went ahead and fixed their problem expecting you to deal with it or shove off Does all that fit better with what you see before you? [All of this is what I've inferred over months, it's my opinion in my words. You won't find this description with anyone else's name attached :-) ] -- Alan McKinnon alan.mckin...@gmail.com
[gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
Trying to migrate from Ubuntu 10.04 LTS, I have installed Gentoo, profile AMD64 13.0 desktop gnome. Everything so far works as expected except for Internet browsers and GNOME applets that should connect to the Internet. More exactly, Internet browsers (I have tried Firefox and Links) can connect to the WWW if I type the IP address as, for example, 173.194.71.104 but cannot if I type www.google.com And this is strange because the host utility works as expected (converting www.google.com into 173.194.71.104) just adding the following Warning: query response not set. The ifconfig utility shows that instead of eth0 I have somewhat strange enp2s15 but I have played with this and found no difference after forcing udev to name it eth0. Installing Gentoo, I closely followed the instructions from Gentoo Linux AMD64 Handbook (http://www.gentoo.org/doc/en/handbook/handbook-amd64.xml?full=1). The contents of my /etc/conf.d/net file is like this: hostname=myhostname dns_domain_lo=mynetwork config_enp2s15=192.168.0.9 netmask 255.255.255.0 broadcast 192.168.0.255 routes_enp2s15=default via my_rooter_local_network_IP dns_servers_enp2s15=my_rooter_local_network_IP my_provider_DNS_server_IP another_DNS_server_IP The contents of my /etc/resolv.conf is like this: nameserver my_rooter_local_network_IP nameserver my_provider_DNS_server_IP nameserver another_DNS_server_IP The relevant part of /var/log/rc.log is like this: * Bringing up interface enp2s15 * Skipping module adsl due to missing program: /usr/sbin/adsl-start /usr/sbin/pppoe-start * Skipping module br2684ctl due to missing program: br2684ctl * Skipping module tuntap due to missing program: ip openvpn tunctl * Skipping module bridge due to missing program: brctl * Skipping module clip due to missing program: /usr/sbin/atmsigd * Skipping module ethtool due to missing program: ethtool * Skipping module macvlan due to missing program: ip * Skipping module netplugd due to missing program: /sbin/netplugd * Skipping module ifplugd due to missing program: /usr/sbin/ifplugd * Skipping module ipppd due to missing program: /usr/sbin/ipppd * Skipping module iwconfig due to missing program: /sbin/iwconfig * Skipping module iproute2 due to missing program: ip * Skipping module firewalld due to missing program: firewall-cmd * Skipping module pppd due to missing program: /usr/sbin/pppd * Skipping module vlan due to missing program: ip * Skipping module dhclient due to missing program: /sbin/dhclient * Skipping module pump due to missing program: /sbin/pump * Skipping module ip6rd due to missing program: ip * Skipping module ip6to4 due to missing program: ip * Loaded modules: apipa arping bonding ccwgroup macchanger macnet wpa_supplicant ssidnet ifconfig system dhcpcd * Configuring enp2s15 for MAC address my_ethernet_card_mac_address ... [ ok ] * 192.168.0.9 ... [ ok ] * Adding routes * default via my_rooter_local_network_IP ... Remark: My router works ok for the Ubuntu installation on the same computer. Networking options of my kernel is as follows: CONFIG_PACKET=y CONFIG_PACKET_DIAG=y CONFIG_UNIX=y CONFIG_UNIX_DIAG=y CONFIG_XFRM=y CONFIG_XFRM_ALGO=y CONFIG_XFRM_USER=y CONFIG_INET=y CONFIG_INET_TUNNEL=y CONFIG_INET_LRO=y CONFIG_INET_DIAG=y CONFIG_INET_TCP_DIAG=y CONFIG_INET_UDP_DIAG=y CONFIG_TCP_CONG_ADVANCED=y CONFIG_TCP_CONG_CUBIC=y CONFIG_DEFAULT_CUBIC=y CONFIG_DEFAULT_TCP_CONG=cubic CONFIG_TCP_MD5SIG=y CONFIG_IPV6=y CONFIG_INET6_AH=y CONFIG_INET6_ESP=y CONFIG_INET6_XFRM_MODE_TRANSPORT=y CONFIG_INET6_XFRM_MODE_TUNNEL=y CONFIG_INET6_XFRM_MODE_BEET=y CONFIG_IPV6_SIT=y CONFIG_IPV6_NDISC_NODETYPE=y CONFIG_NETWORK_SECMARK=y CONFIG_HAVE_NET_DSA=y CONFIG_DNS_RESOLVER=y CONFIG_RPS=y CONFIG_RFS_ACCEL=y CONFIG_XPS=y CONFIG_BQL=y where all options that are not set has been deleted for brevity. I have also tried to use dhcpd instead of static IP but it did not help. Currently, I have completely exhausted all my creativity trying to figure out how to make my WWW browsers to work properly. Please, help!
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
More exactly, Internet browsers (I have tried Firefox and Links) can connect to the WWW if I type the IP address as, for example, 173.194.71.104 but cannot if I type www.google.com And this is strange because the host utility works as expected (converting www.google.com into 173.194.71.104) just adding the following Warning: query response not set. Do other command line tools like ping work? It looks like your DNS setup is correct but for some reason the GNOME setup isn't. Few things: - Are you using NetworkManager? Is the 'networkmanager' USE flag enabled? If so try going to Settings-Network-Wired-gear icon-IPv4 and see what the DNS settings are like. - Also in Settings-Network, see if Network proxy is (mis)configured. --Mark
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
On Sunday 04 Aug 2013 19:56:08 gevisz wrote: Trying to migrate from Ubuntu 10.04 LTS, I have installed Gentoo, profile AMD64 13.0 desktop gnome. Everything so far works as expected except for Internet browsers and GNOME applets that should connect to the Internet. More exactly, Internet browsers (I have tried Firefox and Links) can connect to the WWW if I type the IP address as, for example, 173.194.71.104 but cannot if I type www.google.com And this is strange because the host utility works as expected (converting www.google.com into 173.194.71.104) just adding the following Warning: query response not set. Hmm .. wonder if this is something to do with your dns repeater (your lan router) sending malformed packets in response? The ifconfig utility shows that instead of eth0 I have somewhat strange enp2s15 but I have played with this and found no difference after forcing udev to name it eth0. This is the new kernel naming scheme of NICs. Which-ever nomenclature you decide to use, check that that's the only one having a symlink in /etc/init.d/ to net.lo There's been a few threads in this list a couple of months ago or so. You may want to refer to them. Installing Gentoo, I closely followed the instructions from Gentoo Linux AMD64 Handbook (http://www.gentoo.org/doc/en/handbook/handbook-amd64.xml?full=1). The contents of my /etc/conf.d/net file is like this: hostname=myhostname dns_domain_lo=mynetwork config_enp2s15=192.168.0.9 netmask 255.255.255.0 broadcast 192.168.0.255 routes_enp2s15=default via my_rooter_local_network_IP dns_servers_enp2s15=my_rooter_local_network_IP my_provider_DNS_server_IP another_DNS_server_IP If you have concluded on using eth0, then change the above entries accordingly and also try leaving only your router's local_network_IP in there. The router should provide you with your ISPs repeaters addresses anyway. The contents of my /etc/resolv.conf is like this: nameserver my_rooter_local_network_IP nameserver my_provider_DNS_server_IP nameserver another_DNS_server_IP The relevant part of /var/log/rc.log is like this: * Bringing up interface enp2s15 * Skipping module adsl due to missing program: /usr/sbin/adsl-start /usr/sbin/pppoe-start * Skipping module br2684ctl due to missing program: br2684ctl * Skipping module tuntap due to missing program: ip openvpn tunctl * Skipping module bridge due to missing program: brctl * Skipping module clip due to missing program: /usr/sbin/atmsigd * Skipping module ethtool due to missing program: ethtool * Skipping module macvlan due to missing program: ip * Skipping module netplugd due to missing program: /sbin/netplugd * Skipping module ifplugd due to missing program: /usr/sbin/ifplugd * Skipping module ipppd due to missing program: /usr/sbin/ipppd * Skipping module iwconfig due to missing program: /sbin/iwconfig * Skipping module iproute2 due to missing program: ip * Skipping module firewalld due to missing program: firewall-cmd * Skipping module pppd due to missing program: /usr/sbin/pppd * Skipping module vlan due to missing program: ip * Skipping module dhclient due to missing program: /sbin/dhclient * Skipping module pump due to missing program: /sbin/pump * Skipping module ip6rd due to missing program: ip * Skipping module ip6to4 due to missing program: ip * Loaded modules: apipa arping bonding ccwgroup macchanger macnet You seem to have built a tonne of modules which you probably do not need. wpa_supplicant ssidnet ifconfig system dhcpcd * Configuring enp2s15 for MAC address my_ethernet_card_mac_address ... [ ok ] * 192.168.0.9 ... [ ok ] * Adding routes * default via my_rooter_local_network_IP ... Remark: My router works ok for the Ubuntu installation on the same computer. Networking options of my kernel is as follows: CONFIG_PACKET=y CONFIG_PACKET_DIAG=y CONFIG_UNIX=y CONFIG_UNIX_DIAG=y CONFIG_XFRM=y CONFIG_XFRM_ALGO=y CONFIG_XFRM_USER=y CONFIG_INET=y CONFIG_INET_TUNNEL=y CONFIG_INET_LRO=y CONFIG_INET_DIAG=y CONFIG_INET_TCP_DIAG=y CONFIG_INET_UDP_DIAG=y CONFIG_TCP_CONG_ADVANCED=y CONFIG_TCP_CONG_CUBIC=y CONFIG_DEFAULT_CUBIC=y CONFIG_DEFAULT_TCP_CONG=cubic CONFIG_TCP_MD5SIG=y CONFIG_IPV6=y CONFIG_INET6_AH=y CONFIG_INET6_ESP=y CONFIG_INET6_XFRM_MODE_TRANSPORT=y CONFIG_INET6_XFRM_MODE_TUNNEL=y CONFIG_INET6_XFRM_MODE_BEET=y CONFIG_IPV6_SIT=y CONFIG_IPV6_NDISC_NODETYPE=y CONFIG_NETWORK_SECMARK=y CONFIG_HAVE_NET_DSA=y CONFIG_DNS_RESOLVER=y CONFIG_RPS=y CONFIG_RFS_ACCEL=y CONFIG_XPS=y CONFIG_BQL=y where all options that are not set has been deleted for brevity. I have also tried to use dhcpd instead of static IP but it did not help. Currently, I have completely exhausted all my creativity trying to figure out how to make my WWW browsers to work properly. Please, help! Can you please show us: ip route show ip addr show ip link show -- Regards, Mick signature.asc Description: This is a digitally signed
Re: [gentoo-user] Browsers cannot access WWW while ping and host utilities work as expected.
Can you do a ping and see if the resolv.conf DNS ips are reachable? do a dig @8.8.8.8 www.google.com ## which will do a name resolution with Google DNS servers. also pls post make sure you have correct routing table (route -n) Regards, Kurian. On Sun, Aug 4, 2013 at 7:57 PM, Mick michaelkintz...@gmail.com wrote: On Sunday 04 Aug 2013 19:56:08 gevisz wrote: Trying to migrate from Ubuntu 10.04 LTS, I have installed Gentoo, profile AMD64 13.0 desktop gnome. Everything so far works as expected except for Internet browsers and GNOME applets that should connect to the Internet. More exactly, Internet browsers (I have tried Firefox and Links) can connect to the WWW if I type the IP address as, for example, 173.194.71.104 but cannot if I type www.google.com And this is strange because the host utility works as expected (converting www.google.com into 173.194.71.104) just adding the following Warning: query response not set. Hmm .. wonder if this is something to do with your dns repeater (your lan router) sending malformed packets in response? The ifconfig utility shows that instead of eth0 I have somewhat strange enp2s15 but I have played with this and found no difference after forcing udev to name it eth0. This is the new kernel naming scheme of NICs. Which-ever nomenclature you decide to use, check that that's the only one having a symlink in /etc/init.d/ to net.lo There's been a few threads in this list a couple of months ago or so. You may want to refer to them. Installing Gentoo, I closely followed the instructions from Gentoo Linux AMD64 Handbook (http://www.gentoo.org/doc/en/handbook/handbook-amd64.xml?full=1). The contents of my /etc/conf.d/net file is like this: hostname=myhostname dns_domain_lo=mynetwork config_enp2s15=192.168.0.9 netmask 255.255.255.0 broadcast 192.168.0.255 routes_enp2s15=default via my_rooter_local_network_IP dns_servers_enp2s15=my_rooter_local_network_IP my_provider_DNS_server_IP another_DNS_server_IP If you have concluded on using eth0, then change the above entries accordingly and also try leaving only your router's local_network_IP in there. The router should provide you with your ISPs repeaters addresses anyway. The contents of my /etc/resolv.conf is like this: nameserver my_rooter_local_network_IP nameserver my_provider_DNS_server_IP nameserver another_DNS_server_IP The relevant part of /var/log/rc.log is like this: * Bringing up interface enp2s15 * Skipping module adsl due to missing program: /usr/sbin/adsl-start /usr/sbin/pppoe-start * Skipping module br2684ctl due to missing program: br2684ctl * Skipping module tuntap due to missing program: ip openvpn tunctl * Skipping module bridge due to missing program: brctl * Skipping module clip due to missing program: /usr/sbin/atmsigd * Skipping module ethtool due to missing program: ethtool * Skipping module macvlan due to missing program: ip * Skipping module netplugd due to missing program: /sbin/netplugd * Skipping module ifplugd due to missing program: /usr/sbin/ifplugd * Skipping module ipppd due to missing program: /usr/sbin/ipppd * Skipping module iwconfig due to missing program: /sbin/iwconfig * Skipping module iproute2 due to missing program: ip * Skipping module firewalld due to missing program: firewall-cmd * Skipping module pppd due to missing program: /usr/sbin/pppd * Skipping module vlan due to missing program: ip * Skipping module dhclient due to missing program: /sbin/dhclient * Skipping module pump due to missing program: /sbin/pump * Skipping module ip6rd due to missing program: ip * Skipping module ip6to4 due to missing program: ip * Loaded modules: apipa arping bonding ccwgroup macchanger macnet You seem to have built a tonne of modules which you probably do not need. wpa_supplicant ssidnet ifconfig system dhcpcd * Configuring enp2s15 for MAC address my_ethernet_card_mac_address ... [ ok ] * 192.168.0.9 ... [ ok ] * Adding routes * default via my_rooter_local_network_IP ... Remark: My router works ok for the Ubuntu installation on the same computer. Networking options of my kernel is as follows: CONFIG_PACKET=y CONFIG_PACKET_DIAG=y CONFIG_UNIX=y CONFIG_UNIX_DIAG=y CONFIG_XFRM=y CONFIG_XFRM_ALGO=y CONFIG_XFRM_USER=y CONFIG_INET=y CONFIG_INET_TUNNEL=y CONFIG_INET_LRO=y CONFIG_INET_DIAG=y CONFIG_INET_TCP_DIAG=y CONFIG_INET_UDP_DIAG=y CONFIG_TCP_CONG_ADVANCED=y CONFIG_TCP_CONG_CUBIC=y CONFIG_DEFAULT_CUBIC=y CONFIG_DEFAULT_TCP_CONG=cubic CONFIG_TCP_MD5SIG=y CONFIG_IPV6=y CONFIG_INET6_AH=y CONFIG_INET6_ESP=y CONFIG_INET6_XFRM_MODE_TRANSPORT=y CONFIG_INET6_XFRM_MODE_TUNNEL=y CONFIG_INET6_XFRM_MODE_BEET=y CONFIG_IPV6_SIT=y CONFIG_IPV6_NDISC_NODETYPE=y CONFIG_NETWORK_SECMARK=y CONFIG_HAVE_NET_DSA=y CONFIG_DNS_RESOLVER=y CONFIG_RPS=y CONFIG_RFS_ACCEL=y CONFIG_XPS=y CONFIG_BQL=y
