| metala | Q: I am having issue with ifupdown and wg-quick. Sometime I come up with Wireguard up, but with no route to the VPN endpoint. I've tried to add preup script, but I cannot determine what is the main device (eth0,wlan0), then I've tried script in ifup.d/ but the wlan0 is up before it gets associated and therefore ip route command fails. What would be your approach? | 07:37 |
|---|---|---|
| rrq | if wlan0 is brought up using the "default" stanza then an "up" command for that might work... or is that what you have tried? | 07:52 |
| metala | I manually do # ifup wlan0 | 07:56 |
| rrq | yes, but wp_suulicant uses wpa_action which uses ifup when wlan0 gets connected | 07:58 |
| rrq | wpa_supplicant | 07:58 |
| metala | /etc/network/interfaces: iface wlan0 inet manual; wpa-roam /etc/wpa_supplicant/wpa-roam.conf // I depend on wpa_gui | 07:58 |
| rrq | no difference... wpa_gui merely configures for wpa_supplicant... it's wpa_supplicant that manages the link layer | 07:59 |
| rrq | it ends up using "ifup default" | 07:59 |
| rrq | (unless there an id for the access point) | 08:00 |
| rrq | id_string I think it's called.. then it'd use idup $id_string | 08:00 |
| rrq | ifup $id_string | 08:01 |
| metala | I might need to read a bit more about this, I guess I am lacking some knowledge there. | 08:01 |
| rrq | (I've never used that aspect) | 08:01 |
| metala | I just wanted on post up to add a static route to the interface that gets brought up | 08:02 |
| rrq | i thought you wanted to do something when the interface gets connected | 08:03 |
| metala | it is. I want add a static route on eth0 link up (which I do manually with ifup) and on wlan0 after association and dhcp lease acquisition. | 08:05 |
| rrq | right; wp_supplicant ends up invoking "ifup default" when the wireless link gets connected | 08:06 |
| rrq | normally you'd have an iface block for thatm like "iface default inet dhcp" | 08:07 |
| metala | are you sure? > # ifup default; Cannot find device "default" | 08:07 |
| metala | ah... I need to define a default interface. I see | 08:08 |
| rrq | check "man wpa_action" and "man wpa_supplicant" | 08:09 |
| rrq | and "man wpa_gui" as well I suppose | 08:09 |
| metala | thank you | 08:09 |
| rrq | nw | 08:10 |
| rrq | ... and then I think some dhcp clients also have connection-up hooks | 08:13 |
| rrq | I'm using udhcpc myself | 08:13 |
| metala | the eth0 is manual, that's why I wanted to do it with an ifup script. | 08:19 |
| metala | but I guess I can make it generic and reference it from both | 08:22 |
| taylan | Yesterday I pondered how much longer the T61 I just set up will survive; this morning I woke up to it being totally unresponsive (including over SSH) with a black screen, and had to hold the power button to shut it down and restart it... | 13:09 |
| taylan | It's running fine again now. Might there be a way to find out what happened? Like, a kernel panic that may have been logged somewhere? | 13:09 |
| ama | Hello! Is Gluster working on Devuan? I've installed the gluster-cli, -client, -common, and -server packages but there seem to be no Gluster service working for me. The 'gluster version' command tells me I'm running glusterfs 10.3 (Copyright 2006-2016). Commands to probe peers say: 'Connection failed. Please check if gluster daemon is operational.' Any ideas please? | 13:18 |
| rustyaxe | taylan: usually kernel panic wont write anything to disk, because at that point the kernel is known to be in a damaged state. does it have a serial port/ If so you could hook it to another machine and set things up to use that as a serial console (edit grub config) -- then the panic will can be logged on other machine | 14:27 |
| taylan | hmm I see. for now, I've disabled all the services I don't need, like slim, cups, saned, avahi, elogind, dbus, and network-manager (installed dhcpcd instead) and will hope it doesn't happen again, but maybe I should run a memtest or so at some point since the device is ancient | 14:31 |
| ama | I've been told on #Gluster that GlusterFS doesn't deppend on systemd, but the packages on Devuan (daedalus) are probably missing the init.d scripts. Is there a place I could get those scripts from an older (pre systemd rubish) Debian release or something like that, please? Any ideas? | 14:34 |
| ama | I've copied the /etc/init.d/gluster-server script from a Debian 7 (pre systemd) into my Devuan daedalus and now: 'glusterd service is running with pid 29910.'. It seems it's working! | 14:46 |
| hiddener | ama: yay! | 14:46 |
| hiddener | one day you will write service scripts like a pro :D | 14:47 |
| ama | apt show glusterfs-server shows: APT-Sources: http://deb.devuan.org/merged daedalus/main amd64 Packages... Does that mean that the package is coming from a Devuan repo or could it be teken from a Debian one throug that URL? If the former, maybe the init.d script should/could be included in the package? | 14:48 |
| golinux | Mayber this will help? https://pkginfo.devuan.org/cgi-bin/policy-query.html?c=package&q=gluster*&x=submit | 16:58 |
| golinux | Devuan pulls Debian packages directly. | 16:59 |
| golinux | Amprolla does the magic: https://git.devuan.org/devuan/amprolla3 | 17:01 |
| golinux | See: https://dev1galaxy.org/viewtopic.php?id=3192 | 17:02 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!