Re: [Server-devel] XS 0.5 Second Attempt at eth1
2008/12/5 Reuben K. Caron [EMAIL PROTECTED]: After doing a clean install 4 times and dhcpd not starting Thanks for the dedication and the report - I managed to repro the issue as you describe below. ... When I cd /etc/sysconfig/olpc-scripts ./domain_config xs5.org It creates the xs_domain_name file in /etc/sysconfig/olpc-scripts/ that's the wrong landing place, as you recognised. The current working directory should not affect us. The patch is here: http://dev.laptop.org/git?p=projects/xs-config;a=commitdiff;h=ae030b75d85fa46962134424667dfa6fb7eba5cf and there's an xs-config 0.5.8-biglonghash.rpm on its way to the xs-dev server. Not quite in the same league as the bonding issue, but a good thing to spot and fix. cheers, m -- [EMAIL PROTECTED] [EMAIL PROTECTED] -- School Server Architect - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
After doing a clean install 4 times and dhcpd not starting I investigated a bit more and I found: When I cd /etc/sysconfig/olpc-scripts ./domain_config xs5.org It creates the xs_domain_name file in /etc/sysconfig/olpc-scripts/ However, /etc/sysconfig/dhcpd looks for this file in /etc/sysconfig/ When I copy /etc/sysconfig/olpc-scripts/xs_domain_name to /etc/sysconfig/xs_domain_name; then start dhcpd it correctly generates the config files and starts up. For completeness I then deleted /etc/sysconfig/olpc-scripts/xs_domain_name and /etc/sysconfig/xs_domain_name. Then ran: /etc/sysconfig/olpc-scripts/domain_config xs5.org ; this does generate xs_domain_name in the correct /etc/sysconfig directory. Reuben Anna wrote: On Thu, Dec 4, 2008 at 2:00 PM, Reuben K. Caron [EMAIL PROTECTED] wrote: Actually, I just tried a clean install and DHCP doesn't come up. When I check status it continues to tell me that I must run network-config and domain-config before DHPCd can start; even though I have run those repeatedly. Reuben I did the 0.5 clean install on both machines and dhcpd came up on each one after domain_config. I noticed it worked before I ran network_config, too, for whatever reason. [EMAIL PROTECTED] ~]# service dhcpd status /etc / make: `dhcpd-xs.conf' is up to date. / dhcpd (pid 3451) is running... [EMAIL PROTECTED] ~]# service dhcpd status /etc / make: `dhcpd-xs.conf' is up to date. / dhcpd (pid 3176) is running... On the first machine, I had to manually set the hostname back to alabamaxo.org for the time being so jabber would work, but that's another issue for later. Anna ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Now that I have DHCPD running on a clean install. I can report that it still does not hand out IPs. When I assign my laptop ethernet adapter as 172.18.96.1 and 172.18.0.1 I do get an ARP error from the server; however, when I bump it up one address 172.18.0.2 it still does not ping. Martin Langhoff wrote: On Fri, Dec 5, 2008 at 12:06 AM, Jerry Vonau [EMAIL PROTECTED] wrote: Anna wrote: So, what are the repercussions of this? Not sure... Martin? Good sleuthing! Bonding-related errors was the last thing I'd imagined, and with the e1000 driver too -- it's widely used and generally well maintained. Anna, what's your NIC? Ideally, we want to know the marketing make/name/model and what lspci reports for it. Perhaps I can get my hands on the same hw you have. Googling a bit for e1000 and bonding to see if there are any known issues I see quite a bit of traffic back in 2006/2007 with e1000 devices coming up / powering up late and confusing the bond drivers and sometimes bad performance. But nothing current, it looks like it got sorted out. What's your other NIC? If you run xs-swapnics (and swap the cables too), do things work better with the bonding configuration we're shipping? Also - I'm trying to think this through - I am fairly certain that one of the test machines in Wellington has an e1000 and it worked correctly for me (with a crossover cable to the AP). cheers, m -- Reuben K. Caron Country Support Engineer One Laptop per Child Mobile: +1-617-230-3893 [EMAIL PROTECTED] Deployments Support http://wiki.laptop.org/go/Deployments_Support ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
On Fri, Dec 5, 2008 at 6:13 PM, Raúl Gutiérrez Segalés [EMAIL PROTECTED] wrote: I was hit by the same problem last week.. ARP works but no ping nor IP messages seem to get through. I am far away from the server now but I think the NIC was a CNet. I hate to be away from my test machines. Can't do much today on this track, apologies, but I'll definitely get my hands on a couple of test machines and alternative NICs on Monday and test the hell out of this. cheers, m -- [EMAIL PROTECTED] [EMAIL PROTECTED] -- School Server Architect - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
On Fri, Dec 5, 2008 at 12:06 AM, Jerry Vonau [EMAIL PROTECTED] wrote: Anna wrote: So, what are the repercussions of this? Not sure... Martin? Good sleuthing! Bonding-related errors was the last thing I'd imagined, and with the e1000 driver too -- it's widely used and generally well maintained. Anna, what's your NIC? Ideally, we want to know the marketing make/name/model and what lspci reports for it. Perhaps I can get my hands on the same hw you have. Googling a bit for e1000 and bonding to see if there are any known issues I see quite a bit of traffic back in 2006/2007 with e1000 devices coming up / powering up late and confusing the bond drivers and sometimes bad performance. But nothing current, it looks like it got sorted out. What's your other NIC? If you run xs-swapnics (and swap the cables too), do things work better with the bonding configuration we're shipping? Also - I'm trying to think this through - I am fairly certain that one of the test machines in Wellington has an e1000 and it worked correctly for me (with a crossover cable to the AP). cheers, m -- [EMAIL PROTECTED] [EMAIL PROTECTED] -- School Server Architect - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
On Fri, Dec 5, 2008 at 5:04 AM, Martin Langhoff [EMAIL PROTECTED]wrote: Anna, what's your NIC? Ideally, we want to know the marketing make/name/model and what lspci reports for it. Perhaps I can get my hands on the same hw you have. I'm currently testing on two machines. They're both circa 2003 Dells - GX270 P4's. Server A 02:09.0 Ethernet controller: D-Link System Inc RTL8139 Ethernet (rev 10) 02:0c.0 Ethernet controller: Intel Corporation 82540EM Gigabit Ethernet Controller (rev 02) Server B 01:07.0 Ethernet controller: Macronix, Inc. [MXIC] MX987x5 (rev 25) 01:0c.0 Ethernet controller: Intel Corporation 82540EM Gigabit Ethernet Controller (rev 02) Of course the Intel NICs are the onboard ones. I found the other two in my junkyard, so more than likely they're circa 2000. What's your other NIC? If you run xs-swapnics (and swap the cables too), do things work better with the bonding configuration we're shipping? I tried xs-swapnics on both boxes during my first installation attempts, so I know all four have had the different roles at some point. I've moved the D-Link to different slots in Server A when it was assigned the eth1 role. I've tried using different cables; tried with a hub and without; and tried getting an IP with an AP, a Fedora Desktop, and an Ubuntu desktop. I think I've been rather thorough. So based on all that trial and error and messing around, I think I can definitively report that I can not get bonding to work with any of my hardware. I sure hope it's not an issue with the Dell GX270, because our deployment got hold of a couple dozen of them to use as XS's. Anna Schoolfield Birmingham ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
I was hit by the same problem last week.. ARP works but no ping nor IP messages seem to get through. I am far away from the server now but I think the NIC was a CNet. On Vie, 5 de Diciembre de 2008, 12:34 pm, Reuben K. Caron wrote: Now that I have DHCPD running on a clean install. I can report that it still does not hand out IPs. When I assign my laptop ethernet adapter as 172.18.96.1 and 172.18.0.1 I do get an ARP error from the server; however, when I bump it up one address 172.18.0.2 it still does not ping. Martin Langhoff wrote: On Fri, Dec 5, 2008 at 12:06 AM, Jerry Vonau [EMAIL PROTECTED] wrote: Anna wrote: So, what are the repercussions of this? Not sure... Martin? Good sleuthing! Bonding-related errors was the last thing I'd imagined, and with the e1000 driver too -- it's widely used and generally well maintained. Anna, what's your NIC? Ideally, we want to know the marketing make/name/model and what lspci reports for it. Perhaps I can get my hands on the same hw you have. Googling a bit for e1000 and bonding to see if there are any known issues I see quite a bit of traffic back in 2006/2007 with e1000 devices coming up / powering up late and confusing the bond drivers and sometimes bad performance. But nothing current, it looks like it got sorted out. What's your other NIC? If you run xs-swapnics (and swap the cables too), do things work better with the bonding configuration we're shipping? Also - I'm trying to think this through - I am fairly certain that one of the test machines in Wellington has an e1000 and it worked correctly for me (with a crossover cable to the AP). cheers, m -- Reuben K. Caron Country Support Engineer One Laptop per Child Mobile: +1-617-230-3893 [EMAIL PROTECTED] Deployments Support http://wiki.laptop.org/go/Deployments_Support ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel --- Raúl Gutiérrez Segalés +595 981 231 839 ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Reuben K. Caron wrote: Thanks for clarification. Still no go: Dec 4 12:12:24 schoolserver1 kernel: eth0: link up, 100Mbps, full-duplex, lpa 0xC1E1 Dec 4 12:12:24 schoolserver1 kernel: eth0: link up, 100Mbps, full-duplex, lpa 0xC1E1 Dec 4 12:12:24 schoolserver1 kernel: Ethernet Channel Bonding Driver: v3.3.0 (June 10, 2008) Dec 4 12:12:24 schoolserver1 kernel: bonding: ARP monitoring set to 1000 ms, validate none, with 1 target(s): 172.18.0.1 172.18.0.1 is on the server, no? Think you need to use something that is connected to the wire. Jerry ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Yes, it is on the server. However isn't it a bit chicken and egg setting an ip address that will be on the wire given that the device on the wire is waiting for an ip address from the server? Jerry Vonau wrote: Reuben K. Caron wrote: Thanks for clarification. Still no go: Dec 4 12:12:24 schoolserver1 kernel: eth0: link up, 100Mbps, full-duplex, lpa 0xC1E1 Dec 4 12:12:24 schoolserver1 kernel: eth0: link up, 100Mbps, full-duplex, lpa 0xC1E1 Dec 4 12:12:24 schoolserver1 kernel: Ethernet Channel Bonding Driver: v3.3.0 (June 10, 2008) Dec 4 12:12:24 schoolserver1 kernel: bonding: ARP monitoring set to 1000 ms, validate none, with 1 target(s): 172.18.0.1 172.18.0.1 is on the server, no? Think you need to use something that is connected to the wire. Jerry ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
On Thu, Dec 4, 2008 at 1:14 PM, Reuben K. Caron [EMAIL PROTECTED] wrote: I too am having problems with DHCP (and ejabberd) on XS 0.5. That's interesting! Thanks for the writeup... I reviewed /etc/sysconfig/olpc-scripts/dhcpd.conf.1 - It did not match dhcpd-xs.conf I ran ./network-config 1 I ran ./domain-config xs4.org (i know it isn't xs4 but I'm maintaining domain for posterity) *yes* - you'll notice in the release notes... a special section that talks about upgrading, and asks you to run those commands ;-) Rebooted. dhcpd-xs.conf now has the declaration for 172.18.96 DHCP is started but still does not hand out an IP. That's weird. - any further error msgs on the logs? - maybe the XS is connected to the AP with a straight cable? You might need a crossover... - if you connect a laptop to it via a crossover, can it get a dhcp lease? - maybe the AP is doing NAT? - cheers, m -- [EMAIL PROTECTED] [EMAIL PROTECTED] -- School Server Architect - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Martin Langhoff wrote: On Thu, Dec 4, 2008 at 1:14 PM, Reuben K. Caron [EMAIL PROTECTED] wrote: I too am having problems with DHCP (and ejabberd) on XS 0.5. That's interesting! Thanks for the writeup... I reviewed /etc/sysconfig/olpc-scripts/dhcpd.conf.1 - It did not match dhcpd-xs.conf I ran ./network-config 1 I ran ./domain-config xs4.org (i know it isn't xs4 but I'm maintaining domain for posterity) *yes* - you'll notice in the release notes... a special section that talks about upgrading, and asks you to run those commands ;-) http://wiki.laptop.org/go/XS_Installing_Software#Upgrading_from_XS_0.4_or_earlier Does not include network-config. Should it? Rebooted. dhcpd-xs.conf now has the declaration for 172.18.96 DHCP is started but still does not hand out an IP. That's weird. - any further error msgs on the logs? - maybe the XS is connected to the AP with a straight cable? You might need a crossover... - if you connect a laptop to it via a crossover, can it get a dhcp lease? - maybe the AP is doing NAT? - XS Lan connected to Switch. Bridge AP connected to switch does not pass an IP. Computer connected to Switch does not get an IP. ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
On Thu, Dec 4, 2008 at 4:51 PM, Reuben K. Caron [EMAIL PROTECTED] wrote: http://wiki.laptop.org/go/XS_Installing_Software#Upgrading_from_XS_0.4_or_earlier Does not include network-config. Should it? As far as I tested... no. But I could be wrong about this. Between 4 and 5, the %post install script of xs-config - renames the config file holding the number - re-runs network_config which re-creates any files that we need to maintain XS Lan connected to Switch. Bridge AP connected to switch does not pass an IP. Computer connected to Switch does not get an IP. Hmmm. Right now I don't have a machine handy to test (I'll get my hands on one tomorrow) but I'd love to get a packet capture of eth1 there. So this problem turns up with an upgrade from .4 to .5. Clean 0.5 installs work better for you? cheers, m -- [EMAIL PROTECTED] [EMAIL PROTECTED] -- School Server Architect - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Actually, I just tried a clean install and DHCP doesn't come up. When I check status it continues to tell me that I must run network-config and domain-config before DHPCd can start; even though I have run those repeatedly. Reuben Martin Langhoff wrote: Clean 0.5 installs work better for you? cheers, m ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
On Thu, Dec 4, 2008 at 2:00 PM, Reuben K. Caron [EMAIL PROTECTED] wrote: Actually, I just tried a clean install and DHCP doesn't come up. When I check status it continues to tell me that I must run network-config and domain-config before DHPCd can start; even though I have run those repeatedly. Reuben I did the 0.5 clean install on both machines and dhcpd came up on each one after domain_config. I noticed it worked before I ran network_config, too, for whatever reason. [EMAIL PROTECTED] ~]# service dhcpd status /etc / make: `dhcpd-xs.conf' is up to date. / dhcpd (pid 3451) is running... [EMAIL PROTECTED] ~]# service dhcpd status /etc / make: `dhcpd-xs.conf' is up to date. / dhcpd (pid 3176) is running... On the first machine, I had to manually set the hostname back to alabamaxo.org for the time being so jabber would work, but that's another issue for later. Anna ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Anna wrote: Per Jerry's suggestion, I added this to /etc/modprobe.d/xs_bonding and deleted stuff from the prior suggestion, though I can put that back in if it's going to make a difference. options lanbond0 mode=active-backup arp_interval=1000 arp_ip_target= 172.18.96.5 I went ahead and rebooted. I connected my AP, which is set to that IP address, directly to eth1. Still couldn't ping it. When I do ifdown eth1 then ifup eth1, here's the output from /var/log/messages: Dec 4 12:30:39 schoolserver1 avahi-daemon[3209]: Withdrawing address record for fe80::20f:1fff:fedd:3b93 on eth1. Dec 4 12:30:40 schoolserver1 ntpd[3109]: Deleting interface #14 eth1, fe80::20f:1fff:fedd:3b93#123, interface stats: received=0, sent=0, dropped=0, active_time=951 secs Dec 4 12:30:44 schoolserver1 kernel: ADDRCONF(NETDEV_UP): eth1: link is not ready I even set my Ubuntu machine to 172.18.96.5 netmask 255.255.224.0 and plugged it in directly to the XS. Still nothing. Here's the output of the other stuff you asked for. [EMAIL PROTECTED] ~]# ethtool eth1 Settings for eth1: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Advertised auto-negotiation: Yes Speed: Unknown! (65535) Duplex: Unknown! (255) Port: Twisted Pair PHYAD: 0 Transceiver: internal Auto-negotiation: on Supports Wake-on: umbg Wake-on: g Current message level: 0x0007 (7) Link detected: no No link, not good. [EMAIL PROTECTED] ~]# ip addr 1: lo: LOOPBACK,UP,LOWER_UP mtu 16436 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: eth1: NO-CARRIER,BROADCAST,MULTICAST,UP mtu 1500 qdisc pfifo_fast state DOW N qlen 100 link/ether 00:0f:1f:dd:3b:93 brd ff:ff:ff:ff:ff:ff inet6 fe80::20f:1fff:fedd:3b93/64 scope link valid_lft forever preferred_lft forever 3: eth0: BROADCAST,MULTICAST,UP,LOWER_UP mtu 1500 qdisc pfifo_fast state UNKNO WN qlen 1000 link/ether 00:80:c6:ea:10:2e brd ff:ff:ff:ff:ff:ff inet 192.168.1.9/24 brd 192.168.1.255 scope global eth0 inet6 fe80::280:c6ff:feea:102e/64 scope link valid_lft forever preferred_lft forever 4: bond0: BROADCAST,MULTICAST,MASTER mtu 1500 qdisc noop state DOWN link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff 5: lanbond0: NO-CARRIER,BROADCAST,MULTICAST,MASTER,UP mtu 1500 qdisc noqueue s tate DOWN link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff inet 172.18.96.1/19 brd 172.18.127.255 scope global lanbond0 inet 172.18.1.1/24 brd 172.18.1.255 scope global lanbond0:1 inet 172.18.0.1/24 brd 172.18.0.255 scope global lanbond0:2 again no link. 6: mshbond0: NO-CARRIER,BROADCAST,MULTICAST,MASTER,UP mtu 1500 qdisc noqueue s tate DOWN link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff inet 172.18.10.1/23 brd 172.18.11.255 scope global mshbond0 inet 172.18.16.1/24 brd 172.18.16.255 scope global mshbond0:1 7: mshbond1: NO-CARRIER,BROADCAST,MULTICAST,MASTER,UP mtu 1500 qdisc noqueue s tate DOWN link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff inet 172.18.12.1/23 brd 172.18.13.255 scope global mshbond1 8: mshbond2: NO-CARRIER,BROADCAST,MULTICAST,MASTER,UP mtu 1500 qdisc noqueue s tate DOWN link/ether 00:00:00:00:00:00 brd ff:ff:ff:ff:ff:ff inet 172.18.14.1/23 brd 172.18.15.255 scope global mshbond2 And yes, I know which card is which. If eth0 weren't connected properly to my LAN, I couldn't have scp'd this output over to my main XS 0.5 test box, which has active jabber users on it, so I'm trying to keep from messing with it too much until I know exactly what's going on. On both boxes, I have swapped cables, moved the NICs to different slots, tried connecting with and without hubs, tried connecting to the AP and Fedora and Ubuntu desktops, pretty much everything I can think of. They're both old Dells from circa 2003, though they're different models. Both worked fine with XS 0.4. auto-negotiation appears to be failing, setting the card to a down state. Ok think we need to set the speed of the card by hand, via ifcfg-eth1 file, could you add: ETHTOOL_OPTS= speed 10/100/1000 duplex half/full autoneg off and restart the network. Jerry ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
On Thu, Dec 4, 2008 at 5:26 PM, Jerry Vonau [EMAIL PROTECTED] wrote: auto-negotiation appears to be failing, setting the card to a down state. Ok think we need to set the speed of the card by hand, via ifcfg-eth1 file, could you add: ETHTOOL_OPTS= speed 10/100/1000 duplex half/full autoneg off and restart the network. Jerry I edited /etc/sysconfig/network-scripts/ifcfg-eth1 with that at the bottom: # OLPC School server # This file configures the ethernet port connected to local machines (LAN) ### ### NOTE: This file will be _replaced_ in xs-config updates ### DEVICE=eth1 . /etc/sysconfig/xs_network_config if [ foo$XS_LANBOND0_IPADDR != foo ]; then ONBOOT=yes TYPE=Ethernet MASTER=lanbond0 SLAVE=yes NM_CONTROLLED=no fi ETHTOOL_OPTS= speed 10/100/1000 duplex half/full autoneg off But when I restart the network, it throws out this: Bringing up interface lanbond0: ifcfg-eth1: line 16: 10/100/1000: No such file or directory ifcfg-eth1: line 16: 10/100/1000: No such file or directory ifcfg-eth1: line 16: 10/100/1000: No such file or directory ifcfg-eth1: line 16: 10/100/1000: No such file or directory ifcfg-eth1: line 16: 10/100/1000: No such file or directory RTNETLINK answers: File exists [ OK ] It looks like autonegotiation is still on: [EMAIL PROTECTED] ~]# ethtool eth1 Settings for eth1: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Advertised auto-negotiation: Yes Speed: Unknown! (65535) Duplex: Unknown! (255) Port: Twisted Pair PHYAD: 0 Transceiver: internal Auto-negotiation: on Supports Wake-on: umbg Wake-on: g Current message level: 0x0007 (7) Link detected: no ___ Server-devel mailing list [EMAIL PROTECTED] http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
2008/12/3 Anna [EMAIL PROTECTED]: Has anyone gotten eth1 on XS 0.5 working? I have, at least on 4 different bits of hw, but I worked on it too much to be a good tester of it. Others have installed successfully. One thing that 0.5 has is that it almost always picks the ordering of the NICs _in reverse_ of what 0.4 did. So perhaps your NIC identities are reversed? You can test changing the ethernet cables around or use xs-swapnics (and then reboot). Other than that, I'm a bit lost. Can you get any output from calling ethtool eth1 ? What's the physical hw? It's also possible (though unlikely) that F9 has bad drivers for your specific hw... cheers, m -- [EMAIL PROTECTED] [EMAIL PROTECTED] -- School Server Architect - ask interesting questions - don't get distracted with shiny stuff - working code first - http://wiki.laptop.org/go/User:Martinlanghoff ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
I too am having problems with DHCP (and ejabberd) on XS 0.5. Let me detail my steps Appended upgradeany- ran default install Restarted NICs were swapped. Ran xs-swapnics Check var/log/messages found (snipet): Dec 4 09:36:58 schoolserver1 dhcpd: Listening on LPF/lanbond0/00:00:00:00:00:00/172.18.96.0/19 Dec 4 09:36:58 schoolserver1 dhcpd: Sending on LPF/lanbond0/00:00:00:00:00:00/172.18.96.0/19 Dec 4 09:06:24 schoolserver1 dhcpd: No subnet declaration for lanbond0 (172.18.96.1). Dec 4 09:06:24 schoolserver1 dhcpd: ** Ignoring requests on lanbond0. If this is not what Dec 4 09:06:24 schoolserver1 dhcpd:you want, please write a subnet declaration Dec 4 09:06:24 schoolserver1 dhcpd:in your dhcpd.conf file for the network segment Dec 4 09:06:24 schoolserver1 dhcpd:to which interface lanbond0 is attached. ** Reviewed dhcpd-xs.conf file: It did not have a declaration for 172.18.96.* I reviewed /etc/sysconfig/olpc-scripts/dhcpd.conf.1 - It did not match dhcpd-xs.conf I ran ./network-config 1 I ran ./domain-config xs4.org (i know it isn't xs4 but I'm maintaining domain for posterity) Rebooted. dhcpd-xs.conf now has the declaration for 172.18.96 DHCP is started but still does not hand out an IP. Below are ethtool outputs: [EMAIL PROTECTED] log]# ethtool eth1 Settings for eth1: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Half 1000baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Half 1000baseT/Full Advertised auto-negotiation: Yes Speed: 100Mb/s Duplex: Full Port: Twisted Pair PHYAD: 1 Transceiver: internal Auto-negotiation: on Supports Wake-on: g Wake-on: d Current message level: 0x00ff (255) Link detected: yes [EMAIL PROTECTED] log]# ethtool eth0 Settings for eth0: Supported ports: [ TP MII ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Advertised auto-negotiation: Yes Speed: 100Mb/s Duplex: Full Port: MII PHYAD: 1 Transceiver: internal Auto-negotiation: on Supports Wake-on: pumbg Wake-on: d Current message level: 0x0001 (1) Link detected: yes Martin Langhoff wrote: 2008/12/3 Anna [EMAIL PROTECTED]: Has anyone gotten eth1 on XS 0.5 working? I have, at least on 4 different bits of hw, but I worked on it too much to be a good tester of it. Others have installed successfully. One thing that 0.5 has is that it almost always picks the ordering of the NICs _in reverse_ of what 0.4 did. So perhaps your NIC identities are reversed? You can test changing the ethernet cables around or use xs-swapnics (and then reboot). Other than that, I'm a bit lost. Can you get any output from calling ethtool eth1 ? What's the physical hw? It's also possible (though unlikely) that F9 has bad drivers for your specific hw... cheers, m ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Anna wrote: I tried the XS 0.5 install on another box to see if the LAN issues on my previous attempt were a hardware problem or something. It's a different model Dell with 2 nics. Here are my notes: Basic install from the CD, text mode, default everything. Boot up, then: /etc/sysconfig/olpc-scripts/domain_config anna.org Reboot /etc/sysconfig/olpc-scripts/network_config 1 Reboot Of course, the hostname was still localhost.localdomain so I edited /etc/sysconfig/network for HOSTNAME=schoolserver1 Reboot. Hostname was now updated to schoolserver1. Eth0 pulled an IP from my main LAN, so I knew which card was which. I connected my AP (set to 172.18.96.5 and netmask 255.255.224.0) but couldn't ping it. I tried connecting to the AP with an XO, but of course it didn't do anything. I even connected my Ubuntu machine to the XS, but it couldn't get an IP and I couldn't ping anything either way. Pulled this from your posted log file: Dec 2 21:00:41 alabamaxo kernel: eth0: link up, 100Mbps, full-duplex, lpa 0x41E1 Dec 2 21:00:41 alabamaxo kernel: Ethernet Channel Bonding Driver: v3.3.0 (June 10, 2008) Dec 2 21:00:41 alabamaxo kernel: bonding: MII link monitoring set to 100 ms Dec 2 21:00:41 alabamaxo kernel: bonding: lanbond0 is being created... Dec 2 21:00:41 alabamaxo kernel: ADDRCONF(NETDEV_UP): lanbond0: link is not ready Dec 2 21:00:41 alabamaxo kernel: e1000: eth1: e1000_watchdog: NIC Link is Up 10 Mbps Half Duplex, Flow Control: None Can you post the output of ip addr and ethtool please. What is eth1 plugged into? hub? switch? AP? Half Duplex doesn't quite look right, unless it really is a 10mbps hub. Think we need pass arp_interval= and arp_ip_target= in place of miimon= in xs_bonding, as there is no Port: MII on the gigbit card, sorry I missed that before. If your make the edit and /sbin/ifdown eth1, unload, reload the e1000 module, that should pickup the change when you /sbin/ifup eth1. Tail the log file to see if landbond0 becomes active. Jerry ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Reuben K. Caron wrote: I too am having problems with DHCP (and ejabberd) on XS 0.5. Let me detail my steps Appended upgradeany- ran default install Restarted NICs were swapped. Ran xs-swapnics Check var/log/messages found (snipet): Dec 4 09:36:58 schoolserver1 dhcpd: Listening on LPF/lanbond0/00:00:00:00:00:00/172.18.96.0/19 Dec 4 09:36:58 schoolserver1 dhcpd: Sending on LPF/lanbond0/00:00:00:00:00:00/172.18.96.0/19 Dec 4 09:06:24 schoolserver1 dhcpd: No subnet declaration for lanbond0 (172.18.96.1). Dec 4 09:06:24 schoolserver1 dhcpd: ** Ignoring requests on lanbond0. If this is not what Dec 4 09:06:24 schoolserver1 dhcpd:you want, please write a subnet declaration Dec 4 09:06:24 schoolserver1 dhcpd:in your dhcpd.conf file for the network segment Dec 4 09:06:24 schoolserver1 dhcpd:to which interface lanbond0 is attached. ** Reviewed dhcpd-xs.conf file: It did not have a declaration for 172.18.96.* I reviewed /etc/sysconfig/olpc-scripts/dhcpd.conf.1 - It did not match dhcpd-xs.conf I ran ./network-config 1 I ran ./domain-config xs4.org (i know it isn't xs4 but I'm maintaining domain for posterity) Rebooted. dhcpd-xs.conf now has the declaration for 172.18.96 DHCP is started but still does not hand out an IP. Below are ethtool outputs: [EMAIL PROTECTED] log]# ethtool eth1 Settings for eth1: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Half 1000baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Half 1000baseT/Full Advertised auto-negotiation: Yes Speed: 100Mb/s Duplex: Full Port: Twisted Pair PHYAD: 1 Transceiver: internal Auto-negotiation: on Supports Wake-on: g Wake-on: d Current message level: 0x00ff (255) Link detected: yes OK, is starting to look like the 1000baseT nics are having issues with bonding, can you try the fix I proposed to Anna. Jerry ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Jerry Vonau wrote: Reuben K. Caron wrote: I too am having problems with DHCP (and ejabberd) on XS 0.5. Let me detail my steps Appended upgradeany- ran default install Restarted NICs were swapped. Ran xs-swapnics Check var/log/messages found (snipet): Dec 4 09:36:58 schoolserver1 dhcpd: Listening on LPF/lanbond0/00:00:00:00:00:00/172.18.96.0/19 Dec 4 09:36:58 schoolserver1 dhcpd: Sending on LPF/lanbond0/00:00:00:00:00:00/172.18.96.0/19 Dec 4 09:06:24 schoolserver1 dhcpd: No subnet declaration for lanbond0 (172.18.96.1). Dec 4 09:06:24 schoolserver1 dhcpd: ** Ignoring requests on lanbond0. If this is not what Dec 4 09:06:24 schoolserver1 dhcpd:you want, please write a subnet declaration Dec 4 09:06:24 schoolserver1 dhcpd:in your dhcpd.conf file for the network segment Dec 4 09:06:24 schoolserver1 dhcpd:to which interface lanbond0 is attached. ** Reviewed dhcpd-xs.conf file: It did not have a declaration for 172.18.96.* I reviewed /etc/sysconfig/olpc-scripts/dhcpd.conf.1 - It did not match dhcpd-xs.conf I ran ./network-config 1 I ran ./domain-config xs4.org (i know it isn't xs4 but I'm maintaining domain for posterity) Rebooted. dhcpd-xs.conf now has the declaration for 172.18.96 DHCP is started but still does not hand out an IP. Below are ethtool outputs: [EMAIL PROTECTED] log]# ethtool eth1 Settings for eth1: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Half 1000baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Half 1000baseT/Full Advertised auto-negotiation: Yes Speed: 100Mb/s Duplex: Full Port: Twisted Pair PHYAD: 1 Transceiver: internal Auto-negotiation: on Supports Wake-on: g Wake-on: d Current message level: 0x00ff (255) Link detected: yes OK, is starting to look like the 1000baseT nics are having issues with bonding, can you try the fix I proposed to Anna. OK, I'm tired, all on one line at the bottom /etc/modprobe.d/xs_bonding: options lanbond0 mode=active-backup arp_interval=1000 arp_ip_target=ip-of-AP Jerry ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Thanks for clarification. Still no go: Dec 4 12:12:24 schoolserver1 kernel: eth0: link up, 100Mbps, full-duplex, lpa 0xC1E1 Dec 4 12:12:24 schoolserver1 kernel: eth0: link up, 100Mbps, full-duplex, lpa 0xC1E1 Dec 4 12:12:24 schoolserver1 kernel: Ethernet Channel Bonding Driver: v3.3.0 (June 10, 2008) Dec 4 12:12:24 schoolserver1 kernel: bonding: ARP monitoring set to 1000 ms, validate none, with 1 target(s): 172.18.0.1 Dec 4 12:12:24 schoolserver1 kernel: bonding: lanbond0 is being created... Dec 4 12:12:24 schoolserver1 kernel: ADDRCONF(NETDEV_UP): lanbond0: link is not ready Dec 4 12:12:24 schoolserver1 kernel: ADDRCONF(NETDEV_UP): eth1: link is not ready Dec 4 12:12:24 schoolserver1 kernel: tg3: eth1: Link is up at 100 Mbps, full duplex. Dec 4 12:12:24 schoolserver1 kernel: tg3: eth1: Flow control is on for TX and on for RX. Dec 4 12:12:24 schoolserver1 kernel: ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready Dec 4 12:12:24 schoolserver1 kernel: bonding: mshbond0 is being created... Dec 4 12:12:24 schoolserver1 kernel: ADDRCONF(NETDEV_UP): mshbond0: link is not ready Dec 4 12:12:24 schoolserver1 kernel: bonding: mshbond1 is being created... Dec 4 12:12:24 schoolserver1 kernel: ADDRCONF(NETDEV_UP): mshbond1: link is not ready Dec 4 12:12:24 schoolserver1 kernel: bonding: mshbond2 is being created... Dec 4 12:12:24 schoolserver1 kernel: ADDRCONF(NETDEV_UP): mshbond2: link is not ready ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Not having done this before, and after getting syntax errors and command not found for speed, I googled and edited /etc/sysconfig/network-scripts/ifcfg-eth1 like so. I figured 100 might be a good place to start. I haven't tried half duplex yet, but I can if you think that would do anything. ETHTOOL_OPTS=speed 100 duplex full autoneg off Restarting the network didn't give any errors. After that, ethtool eth1 gives this: [EMAIL PROTECTED] ~]# ethtool eth1 Settings for eth1: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Supports auto-negotiation: Yes Advertised link modes: Not reported Advertised auto-negotiation: No Speed: Unknown! (65535) Duplex: Unknown! (255) Port: Twisted Pair PHYAD: 0 Transceiver: internal Auto-negotiation: off Supports Wake-on: umbg Wake-on: g Current message level: 0x0007 (7) Link detected: no Of course, still no pinging anything. In fact, I don't have blinky lights on the NIC anymore or a link light on the AP, which I had before. Anna ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
Anna wrote: Not having done this before, and after getting syntax errors and command not found for speed, I googled and edited /etc/sysconfig/network-scripts/ifcfg-eth1 like so. I figured 100 might be a good place to start. I haven't tried half duplex yet, but I can if you think that would do anything. ETHTOOL_OPTS=speed 100 duplex full autoneg off Restarting the network didn't give any errors. After that, ethtool eth1 gives this: [EMAIL PROTECTED] ~]# ethtool eth1 Settings for eth1: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Supports auto-negotiation: Yes Advertised link modes: Not reported Advertised auto-negotiation: No Speed: Unknown! (65535) Duplex: Unknown! (255) Port: Twisted Pair PHYAD: 0 Transceiver: internal Auto-negotiation: off Supports Wake-on: umbg Wake-on: g Current message level: 0x0007 (7) Link detected: no Of course, still no pinging anything. In fact, I don't have blinky lights on the NIC anymore or a link light on the AP, which I had before. Anna OK, I'm out of ideas, lets see if we can configure eth1 without bonding, not sure if this is nic driver issue with bonding, or what... backup the ifcfg-eth1 file, copy ifcfg-landbond0, ifcfg-lanbond0:1 ifcfg-lanbond0:2 and to ifcfg-eth1, ifcfg-eth1:1 ifcfg-eth1:2 respectively. edit the ifcfg-eth1* files changing the device line to match the new name, eth1 eth1:1 eth1:2. Please stop the network before editing the files. Jerry ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
Re: [Server-devel] XS 0.5 Second Attempt at eth1
On Thu, Dec 4, 2008 at 6:29 PM, Jerry Vonau [EMAIL PROTECTED] wrote: OK, I'm out of ideas, lets see if we can configure eth1 without bonding, not sure if this is nic driver issue with bonding, or what... backup the ifcfg-eth1 file, copy ifcfg-landbond0, ifcfg-lanbond0:1 ifcfg-lanbond0:2 and to ifcfg-eth1, ifcfg-eth1:1 ifcfg-eth1:2 respectively. edit the ifcfg-eth1* files changing the device line to match the new name, eth1 eth1:1 eth1:2. Please stop the network before editing the files. Jerry W00t! That did it. I'm emailing from my Ubuntu machine on: eth0 Link encap:Ethernet HWaddr 00:0f:1f:80:0d:ea inet addr:172.18.96.2 Bcast:172.18.127.255 Mask:255.255.224.0 inet6 addr: fe80::20f:1fff:fe80:dea/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:49210 errors:0 dropped:0 overruns:0 frame:0 TX packets:47393 errors:0 dropped:0 overruns:0 carrier:0 collisions:2059 txqueuelen:10 RX bytes:27778611 (26.4 MB) TX bytes:8279042 (7.8 MB) Base address:0xdf40 Memory:feae-feb0 My wifi AP is up and running and an XO pulled 172.18.96.3. Obviously there's a hub in the middle. Oh, here's the latest output of ethtool eth1: [EMAIL PROTECTED] ~]# ethtool eth1 Settings for eth1: Supported ports: [ TP ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full 1000baseT/Full Advertised auto-negotiation: Yes Speed: 10Mb/s Duplex: Half Port: Twisted Pair PHYAD: 0 Transceiver: internal Auto-negotiation: on Supports Wake-on: umbg Wake-on: g Current message level: 0x0007 (7) Link detected: yes So, what are the repercussions of this? Anna ___ Server-devel mailing list Server-devel@lists.laptop.org http://lists.laptop.org/listinfo/server-devel
