Re: Two problems still present in RC3
On Friday, December 09, 2011 10:05:49 am Damien Fleuriot wrote: On 12/9/11 10:13 AM, Brett Glass wrote: FreeBSD 9.0-RC3 is looking good, but I'm still encountering two problems. Firstly, when I try to configure VLANs in /etc/rc.conf, I'm getting errors. For example, if I use vlans_re0=1 2 ip_addrs_re0_1=192.168.0.1-4/16 ip_addrs_re0_2=10.0.0.0/24 to create two VLANs on the interface re0, I get error messages saying that create commands (presumably using ifconfig) have failed. The interfaces SEEM to be configured correctly, but the messages -- which must be coming from scripts called by /etc/netstart -- are troubling. Secondly, there's still some strangeness in the sc terminal emulation. When I run jove, the status line at the bottom of the screen isn't entirely in reverse video as it should be. Only parts of it are, and the highlighting changes -- seemingly at random -- as I work. Neither of these is likely to be a showstopper (so long as the first won't cause me networking problems I haven't observed yet), but both are probably worth looking into. I have never seen this way of configuring VLANs. It's new. I added it in the last year or so to match the way wlan devices are created. It also can do sane things like auto-destroy associated vlans when a device is stopped, etc. However, I have not tested them with /etc/netstart. They do work fine during a normal startup. I can't see obvious reasons as to why they would fail though via /etc/netstart. The stuff that handles vlans_if will auto-load if_vlan.ko if it is needed, but perhaps that isn't working from single user mode when using /etc/netstart? Also, do the vlans work fine if you let a box boot normally? -- John Baldwin ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
On 12/11/11, Brett Glass br...@lariat.net wrote: At 11:36 AM 12/11/2011, Ben Kaduk wrote: Did you take the change to /etc/ttys going from cons25 to xterm 'type'? I didn't have to change it; it was that way when the OS was installed. You were sending to -stable, so I wasn't sure. Thanks for confirming. Problem seems to be that the behavior (specifically, reverse video on the 25th line) doesn't quite match the xterm termcap entry. If I remember correctly, your original message mentioned seeing this issue in emacs; have you tried reproducing it in a simpler test case? Maybe something like: (for i in $(jot 25); do echo; tput mr; echo -n reverse; tput me; echo -n regular; done) sleep 1 That would presumably help narrow down the issue. -Ben Kaduk ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
* Ben Kaduk minimar...@gmail.com, 20111213 20:34: Problem seems to be that the behavior (specifically, reverse video on the 25th line) doesn't quite match the xterm termcap entry. If I remember correctly, your original message mentioned seeing this issue in emacs; have you tried reproducing it in a simpler test case? Well, I don't care whether the test case is small or large, as long as I know how to reproduce it. I have never used jove before, so what kind of buttons should I press to get the artifacts? Does it only happen on the console, or also when using a regular xterm? Thanks, -- Ed Schouten e...@80386.nl WWW: http://80386.nl/ pgpXk4FUWsqI1.pgp Description: PGP signature
Re: Two problems still present in RC3
At 12:34 PM 12/13/2011, Ben Kaduk wrote: If I remember correctly, your original message mentioned seeing this issue in emacs; have you tried reproducing it in a simpler test case? No; when we hit the bug, we moved to SSH with a VT100 emulator so that we could configure the system. But the system console really should work as well. I expect to be configuring another system this evening and will try the script fragment you sent. --Brett ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
At 01:23 PM 12/13/2011, Ed Schouten wrote: Well, I don't care whether the test case is small or large, as long as I know how to reproduce it. I have never used jove before, so what kind of buttons should I press to get the artifacts? Just start inserting text, deleting characters and lines (Ctrl-D and Ctrl-K), etc. It's very obvious. Does it only happen on the console, or also when using a regular xterm? I do not use regular xterms, so I can't answer that, alas. --Brett Glass ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
On Tue, 13 Dec 2011, Brett Glass wrote: At 12:34 PM 12/13/2011, Ben Kaduk wrote: If I remember correctly, your original message mentioned seeing this issue in emacs; have you tried reproducing it in a simpler test case? No; when we hit the bug, we moved to SSH with a VT100 emulator so that we could configure the system. But the system console really should work as well. I expect to be configuring another system this evening and will Definitely! I use the system console pretty regularly, but I just haven't gotten a chance to pull my laptop up to an RC or -current, recently, so I can't test it myself. try the script fragment you sent. Thanks. (It turns out one has to consult terminfo(5) to figure out what attributes to pass in to tput; it might be worth adding an EXAMPLES section to tput(1).) -Ben ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
Hi Brett, * Brett Glass br...@lariat.net, 20111214 03:51: Just start inserting text, deleting characters and lines (Ctrl-D and Ctrl-K), etc. It's very obvious. That's odd. I've just spent some time deleting huge portions of /etc/services and on my system it doesn't seem to do this. Maybe you can do the following: - run script -t 0 ~/log jove ... - Delete characters - As soon as possible as jove starts to misbehave and the screen contents are wrong, run `killall -9 script' - Send me the log Thanks, -- Ed Schouten e...@80386.nl WWW: http://80386.nl/ pgp1bNN1gIMbA.pgp Description: PGP signature
Re: Two problems still present in RC3
Jeremy Chadwick schreef: On Fri, Dec 09, 2011 at 10:30:43AM -0700, Brett Glass wrote: At 03:10 AM 12/9/2011, Johan Hendriks wrote: After a /etc/netstart, i get the following: ifconfig: create: bad value. I get the create: bad value messages as well. What's more, if I change rc.conf to assign variables of the form ifconfig_*, such as ifconfig_re0_1=inet 192.168.0.1 ifconfig_re0_1_alias0=inet 192.168.0.2 ... instead of using an ip_addrs_* variable, I still get those messages. I have confirmed that if I run /etc/netstart after booting, I lose connectivity just as you do. I recognize the challenge of specifying the network configuration in a declaratory rather than a procedural environment, because there are so many contingencies and possible combinations of parameters. But there doesn't seem to be any combination of variables I can assign in rc.conf that doesn't cause errors when I try to create VLANs. I would advise everyone experiencing those messages or any anomalies of this sort to please set rc_debug=yes in /etc/rc.conf and make sure they have serial or firewire console set up. If you don't have some form of remote console, then this is going to be difficult to debug. See rc.conf(5) for what the option does and therefore why serial console will be needed. The information that's needed is what ifconfig commands are being issued on your systems. This cannot be easily determined from rc.conf variables like ifconfig_snakes_in_the_grass=rocks moss, hence the above request. Your means Brett Glass and Johan Hendriks. Finally, please keep in mind the possibility that both of you are experiencing different problems. I have no evidence at this point to support or refute that claim, but it's something that needs to be considered. Ok did some test on the backup machine, i must do it remotely, so no serial access. Lagg0 consist of two interfaces em0 and em1. The machine booted with the following in /etc/rc.conf, and all is fine this way. ifconfig_em0=up ifconfig_em1=up cloned_interfaces=lagg0 ifconfig_lagg0=laggproto lacp laggport em0 laggport em1 192.168.100.1/24 I changed the ip adress from 192.168.100.1 to 192.168.100.2 in /etc/rc.conf, and added rc_debug=yes ifconfig_em0=up ifconfig_em1=up cloned_interfaces=lagg0 ifconfig_lagg0=laggproto lacp laggport em0 laggport em1 192.168.100.2/24 I did a /etc/netstart nas01-bw ~ # /etc/netstart /etc/rc.d/devd: DEBUG: checkyesno: devd_enable is set to YES. devd already running? (pid=1767). /etc/rc.d/hostid: DEBUG: checkyesno: hostid_enable is set to YES. /etc/rc.d/hostid: DEBUG: run_rc_command: doit: hostid_start /etc/rc.d/hostid: DEBUG: checkyesno: rc_startmsgs is set to YES. Setting hostuuid: 9abf4e13-9a77-b31d-9a77-b31d9fbfee13. /etc/rc.d/hostid: DEBUG: checkyesno: rc_startmsgs is set to YES. Setting hostid: 0xf505f6a8. /etc/rc.d/hostname: DEBUG: run_rc_command: doit: hostname_start /etc/rc.d/ipmon: DEBUG: checkyesno: ipmon_enable is set to NO. /etc/rc.d/ipfilter: DEBUG: checkyesno: ipfilter_enable is set to NO. /etc/rc.d/ipnat: DEBUG: checkyesno: ipnat_enable is set to NO. /etc/rc.d/ipfs: DEBUG: checkyesno: ipfs_enable is set to NO. /etc/rc.d/sppp: DEBUG: run_rc_command: doit: sppp_start /etc/rc.d/netif: DEBUG: run_rc_command: doit: network_start ifconfig: create: bad value /etc/rc.d/netif: DEBUG: Cloned: /etc/rc.d/netif: DEBUG: checkyesno: ipv6_activate_all_interfaces is set to NO. /etc/rc.d/netif: DEBUG: checkyesno: ipv6_activate_all_interfaces is set to NO. /etc/rc.d/netif: DEBUG: checkyesno: ipv6_activate_all_interfaces is set to NO. /etc/rc.d/netif: DEBUG: checkyesno: ipv6_activate_all_interfaces is set to NO. /etc/rc.d/netif: DEBUG: checkyesno: ipv6_activate_all_interfaces is set to NO. /etc/rc.d/netif: DEBUG: checkyesno: ipv6_activate_all_interfaces is set to NO. ifconfig: SIOCSLAGGPORT: Device busy /etc/rc.d/netif: DEBUG: checkyesno: ipv6_activate_all_interfaces is set to NO. Starting Network: lo0 em0 em1 bge0 pflog0 pfsync0 lagg0. /etc/rc.d/netif: DEBUG: checkyesno: rc_startmsgs is set to YES. lo0: flags=8049UP,LOOPBACK,RUNNING,MULTICAST metric 0 mtu 16384 options=3RXCSUM,TXCSUM inet6 ::1 prefixlen 128 inet6 fe80::1%lo0 prefixlen 64 scopeid 0xe inet 127.0.0.1 netmask 0xff00 nd6 options=21PERFORMNUD,AUTO_LINKLOCAL em0: flags=8843UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST metric 0 mtu 1500 options=219bRXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,TSO4,WOL_MAGIC ether 00:1b:21:d4:77:fb inet6 fe80::21b:21ff:fed4:77fb%em0 prefixlen 64 scopeid 0x5 nd6 options=29PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL media: Ethernet autoselect (1000baseT full-duplex) status: active em1: flags=8843UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST metric 0 mtu 1500 options=219bRXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,TSO4,WOL_MAGIC ether 00:1b:21:d4:77:fb inet6 fe80::21b:21ff:fed4:749b%em1 prefixlen 64 scopeid 0x6 nd6
Re: Two problems still present in RC3
On Fri, Dec 9, 2011 at 4:13 AM, Brett Glass br...@lariat.net wrote: Secondly, there's still some strangeness in the sc terminal emulation. When I run jove, the status line at the bottom of the screen isn't entirely in reverse video as it should be. Only parts of it are, and the highlighting changes -- seemingly at random -- as I work. Did you take the change to /etc/ttys going from cons25 to xterm 'type'? -Ben Kaduk ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
At 11:36 AM 12/11/2011, Ben Kaduk wrote: Did you take the change to /etc/ttys going from cons25 to xterm 'type'? I didn't have to change it; it was that way when the OS was installed. Problem seems to be that the behavior (specifically, reverse video on the 25th line) doesn't quite match the xterm termcap entry. --Brett ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
Brett Glass schreef: The interfaces SEEM to be configured correctly, but the messages -- which must be coming from scripts called by /etc/netstart -- are troubling. Same thing happens with lagg0 If i use this config, after a reboot all is fine ifconfig_em0=up ifconfig_em1=up cloned_interfaces=lagg0 ifconfig_lagg0=laggproto lacp laggport em0 laggport em1 192.168.100.1/24 After a /etc/netstart, i get the following: ifconfig: create: bad value. devd already running? (pid=1767). Setting hostuuid: 9abf4e13-9a77-b31d-9a77-b31d9fbfee13. Setting hostid: 0xf505f6a8. ifconfig: create: bad value ifconfig: SIOCSLAGGPORT: Device busy Starting Network: lo0 em0 em1 bge0 pflog0 pfsync0 lagg0. snip lagg0: flags=8843UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST metric 0 mtu 1500 options=219bRXCSUM,TXCSUM,VLAN_MTU,VLAN_HWTAGGING,VLAN_HWCSUM,TSO4,WOL_MAGIC ether 00:1b:21:d4:77:fb inet 192.168.100.1 netmask 0xff00 broadcast 192.168.100.255 inet6 fe80::21b:21ff:fed4:77fb%lagg0 prefixlen 64 scopeid 0xf nd6 options=29PERFORMNUD,IFDISABLED,AUTO_LINKLOCAL media: Ethernet autoselect status: active laggproto lacp laggport: em1 flags=0 laggport: em0 flags=0 add net default: gateway 192.168.1.254 To get rid off the bad value notice, i need the following in /etc/rc.conf comment cloned_interfaces and add create to ifconfig. ifconfig_em0=up ifconfig_em1=up #cloned_interfaces=lagg0 ifconfig_lagg0=create laggproto lacp laggport em0 laggport em1 192.168.100.1/24 This way no error, but i do not know if the lagg interface works that way! I got some issues with the setting the ipaddress and so on, i do not remeber it exactly! it happend last week. The machine was a production machine, and can not be offline for too long. So i changed things back to the first config and rebooted the machine. That way everything is fine. But /etc/netstart does not works for the lagg0 interface. regards, Johan Hendriks ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
On 12/9/11 11:10 AM, Johan Hendriks wrote: Brett Glass schreef: The interfaces SEEM to be configured correctly, but the messages -- which must be coming from scripts called by /etc/netstart -- are troubling. Same thing happens with lagg0 If i use this config, after a reboot all is fine ifconfig_em0=up ifconfig_em1=up cloned_interfaces=lagg0 ifconfig_lagg0=laggproto lacp laggport em0 laggport em1 192.168.100.1/24 I notice you're missing inet in your ifconfig_lagg0 line. Try this one instead: ifconfig_lagg0=laggproto lacp laggport em0 laggport em1 inet 192.168.100.1/24 ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
On 12/9/11 10:13 AM, Brett Glass wrote: FreeBSD 9.0-RC3 is looking good, but I'm still encountering two problems. Firstly, when I try to configure VLANs in /etc/rc.conf, I'm getting errors. For example, if I use vlans_re0=1 2 ip_addrs_re0_1=192.168.0.1-4/16 ip_addrs_re0_2=10.0.0.0/24 to create two VLANs on the interface re0, I get error messages saying that create commands (presumably using ifconfig) have failed. The interfaces SEEM to be configured correctly, but the messages -- which must be coming from scripts called by /etc/netstart -- are troubling. Secondly, there's still some strangeness in the sc terminal emulation. When I run jove, the status line at the bottom of the screen isn't entirely in reverse video as it should be. Only parts of it are, and the highlighting changes -- seemingly at random -- as I work. Neither of these is likely to be a showstopper (so long as the first won't cause me networking problems I haven't observed yet), but both are probably worth looking into. I have never seen this way of configuring VLANs. Find below how I set them up on our firewalls, which works like a charm, but on 8.2. You may still want to give it a try on 9.0RC3, that might solve your problem. ### NETWORKING # Configure link aggregation ifconfig_bce0=up ifconfig_bce1=up ifconfig_em0=up ifconfig_lagg0=laggproto failover laggport bce0 laggport bce1 laggport em0 cloned_interfaces=lagg0 vlan14 vlan24 vlan34 carp14 carp24 carp34 # VLAN14 - WAN ifconfig_vlan14=inet [snip] vlan 14 vlandev lagg0 up # VLAN24 - DMZ ifconfig_vlan24=inet 192.168.24.252/24 vlan 24 vlandev lagg0 up # VLAN34 - LAN ifconfig_vlan34=inet 192.168.34.252/24 vlan 34 vlandev lagg0 up # VLAN 611 - VPNs ifconfig_vlan611=inet 10.106.11.252/24 vlan 611 vlandev lagg0 up ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
Damien Fleuriot schreef: On 12/9/11 11:10 AM, Johan Hendriks wrote: Brett Glass schreef: The interfaces SEEM to be configured correctly, but the messages -- which must be coming from scripts called by /etc/netstart -- are troubling. Same thing happens with lagg0 If i use this config, after a reboot all is fine ifconfig_em0=up ifconfig_em1=up cloned_interfaces=lagg0 ifconfig_lagg0=laggproto lacp laggport em0 laggport em1 192.168.100.1/24 I notice you're missing inet in your ifconfig_lagg0 line. Try this one instead: ifconfig_lagg0=laggproto lacp laggport em0 laggport em1 inet 192.168.100.1/24 It should not be the case, at least the handbook does not mentioning it at all. http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/network-aggregation.html I will try it, when i find a reboot window! regards, Johan Hendriks ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
At 03:10 AM 12/9/2011, Johan Hendriks wrote: After a /etc/netstart, i get the following: ifconfig: create: bad value. I get the create: bad value messages as well. What's more, if I change rc.conf to assign variables of the form ifconfig_*, such as ifconfig_re0_1=inet 192.168.0.1 ifconfig_re0_1_alias0=inet 192.168.0.2 ... instead of using an ip_addrs_* variable, I still get those messages. I have confirmed that if I run /etc/netstart after booting, I lose connectivity just as you do. I recognize the challenge of specifying the network configuration in a declaratory rather than a procedural environment, because there are so many contingencies and possible combinations of parameters. But there doesn't seem to be any combination of variables I can assign in rc.conf that doesn't cause errors when I try to create VLANs. --Brett Glass ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
On 9 Dec 2011, at 18:30, Brett Glass br...@lariat.net wrote: At 03:10 AM 12/9/2011, Johan Hendriks wrote: After a /etc/netstart, i get the following: ifconfig: create: bad value. I get the create: bad value messages as well. What's more, if I change rc.conf to assign variables of the form ifconfig_*, such as ifconfig_re0_1=inet 192.168.0.1 ifconfig_re0_1_alias0=inet 192.168.0.2 ... instead of using an ip_addrs_* variable, I still get those messages. I have confirmed that if I run /etc/netstart after booting, I lose connectivity just as you do. I recognize the challenge of specifying the network configuration in a declaratory rather than a procedural environment, because there are so many contingencies and possible combinations of parameters. But there doesn't seem to be any combination of variables I can assign in rc.conf that doesn't cause errors when I try to create VLANs. --Brett Glass Try creating your VLANs as I described on my previous post and see if that helps ?___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org
Re: Two problems still present in RC3
On Fri, Dec 09, 2011 at 10:30:43AM -0700, Brett Glass wrote: At 03:10 AM 12/9/2011, Johan Hendriks wrote: After a /etc/netstart, i get the following: ifconfig: create: bad value. I get the create: bad value messages as well. What's more, if I change rc.conf to assign variables of the form ifconfig_*, such as ifconfig_re0_1=inet 192.168.0.1 ifconfig_re0_1_alias0=inet 192.168.0.2 ... instead of using an ip_addrs_* variable, I still get those messages. I have confirmed that if I run /etc/netstart after booting, I lose connectivity just as you do. I recognize the challenge of specifying the network configuration in a declaratory rather than a procedural environment, because there are so many contingencies and possible combinations of parameters. But there doesn't seem to be any combination of variables I can assign in rc.conf that doesn't cause errors when I try to create VLANs. I would advise everyone experiencing those messages or any anomalies of this sort to please set rc_debug=yes in /etc/rc.conf and make sure they have serial or firewire console set up. If you don't have some form of remote console, then this is going to be difficult to debug. See rc.conf(5) for what the option does and therefore why serial console will be needed. The information that's needed is what ifconfig commands are being issued on your systems. This cannot be easily determined from rc.conf variables like ifconfig_snakes_in_the_grass=rocks moss, hence the above request. Your means Brett Glass and Johan Hendriks. Finally, please keep in mind the possibility that both of you are experiencing different problems. I have no evidence at this point to support or refute that claim, but it's something that needs to be considered. -- | Jeremy Chadwickjdc at parodius.com | | Parodius Networking http://www.parodius.com/ | | UNIX Systems Administrator Mountain View, CA, US | | Making life hard for others since 1977. PGP 4BD6C0CB | ___ freebsd-stable@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to freebsd-stable-unsubscr...@freebsd.org