| FilipZ | Hi! I was here previously asking about help with configuring interfaces file, so the wpa_gui would properly kill the malfunctioning, locked dhclient processes, before trying to connect with the access point again. I tried with adding "pre-up pkill -9 dhclient;", "pre-up pkill dhclient; sleep 1; pkill -9 dhclient", and "down pkill dhclient" after being recommended to do so, but none of these options worked out, and with the first o | 01:46 |
|---|---|---|
| FilipZ | t didn't even turn dhclient process up at the startup. Does anyone have some idea why would it be like this, and how to fix it properly? Also, I don't know why but sometimes when connecting 2 dhclient processes are spawned at once. | 01:46 |
| rrq | Hi FilipZ. note that dhclient and wpa-supplicant are rather separate things and dhclient in general controls itself. | 02:05 |
| rrq | wpa-suplicant manages the link level of the connection, and you may monitor with "tail -F /run/wpa_action.wlan0.timestamp" the times when it start/restarts link level connections | 02:08 |
| rrq | luckily wpa-supplicant performs its actions via wpa_action which is an sh script (that includes /etc/wpa_supplicant/functions.sh) and thus easily accessible wrt what it does | 02:13 |
| rrq | did you try the variant with "down dhclient -x $IFACE" added to the wlan0 stanza? (that command is a different attempt to kill dhclient[s]) | 02:30 |
| rrq | in fact I realized now that all those "kill attempts" should have been on the wlan0 stanza rather than the default stanza | 02:32 |
| rrq | apparently wpa_action performs ifdown only on the physical (wlan0) interface and only ifup on the logical (default) interface | 02:33 |
| FilipZ | I didn't try that approach. What is the -x here? Instead of $IFACE, I should put wlan0, right? | 02:38 |
| rrq | yes $IFACE is wlan0 in that stanza (sorry, my habit to use that variable)... se man dhclient about -x | 02:39 |
| rrq | I don't know which means dhclient implements for "Stop the running DHCP client" .. maybe the command needs to tell about the pid file | 02:43 |
| rrq | ie "-pf /run/dhclient.wlan0.pid" | 02:43 |
| rrq | possibly more important then mentioning "wlan0" | 02:45 |
| FilipZ | rrq: https://paste.debian.net/1258378/ Is it in the proper place like this? | 02:58 |
| FilipZ | After I did "sudo ifdown --force wlan0" it didn't kill the dhclient process, though. | 03:01 |
| rrq | a purist would indent it with 4 spaces .. and maybe it needs -pf /run/dhclient.wlan0.pid | 03:02 |
| FilipZ | In the state of interfaces from that link, that is. | 03:02 |
| rrq | (the indentation is not important) | 03:02 |
| FilipZ | What would "-pf /run/dhclient.wlan0.pid" do, and where should I put it? | 03:03 |
| rrq | down dhclient -x -pf /run/dhclient.wlan0.pid wlan0 | 03:03 |
| rrq | it tells the process wht the pid file is for the previous process | 03:04 |
| FilipZ | This seem to work for now, but what if there will be a several dhclient processes spawned. It will kill only one, right? | 03:06 |
| rrq | probably, perhaps a forecful "down pkill -9 dhclient" as a next step would deal with that? | 03:07 |
| rrq | still in the wlan0 stanza | 03:08 |
| FilipZ | Right. I made it like this. I will write later about whether it worked or not. | 03:09 |
| micdud | shameless plug. https://github.com/dandudikof/zfs-auto-mod my first open sourced project | 03:22 |
| rrq | mmm you know that github is microsoft, yes? | 03:23 |
| FilipZ | The issue occurred again. It fixes itself properly now! :) | 03:30 |
| micdud | rrq, should i be worried about something? | 03:30 |
| rrq | it's fully your choice whether or not to worry about hosting your software with microsoft. | 03:38 |
| Atari-Frosch | Hi. I just tried three Devuan mirrors in order to avoid the pkgmaster, and with all three apt tells me there is no Release file. I went there with a browser, and it's correct: There is none. | 17:15 |
| Atari-Frosch | To sources.list I had written: deb http://dist-mirror.fem.tu-ilmenau.de/devuan chimaera main contrib non-free | 17:16 |
| Atari-Frosch | is that wrong? | 17:16 |
| Atari-Frosch | (same with ungleich.ch and koddos.net) | 17:16 |
| gnarface | did you try using deb.devuan.org? | 17:17 |
| Atari-Frosch | That one works. | 17:17 |
| Atari-Frosch | I had the idea that using mirrors would be a good idea. :-) | 17:17 |
| Atari-Frosch | Got the servers from this list: https://pkgmaster.devuan.org/mirror_list.txt | 17:17 |
| gnarface | i think it's possible you caught those 3 at a bad time, while automated updates were happening | 17:18 |
| gnarface | i'd just wait and try again later to see if it's a real problem | 17:18 |
| Atari-Frosch | Oh, OK. Thank you, I'll try that. | 17:18 |
| gnarface | deb.devuan.org should be a dns round-robin of the ones in the mirror_list.txt | 17:19 |
| Atari-Frosch | Ah! Then I don't have to use the others directly? | 17:19 |
| gnarface | nope, it's advised to use deb.devuan.org | 17:19 |
| Atari-Frosch | thx | 17:19 |
| gnarface | when updates are being pushed, some of them may be temporarily in an unusable state but they try to make sure not to update them all at once | 17:20 |
| Atari-Frosch | I see. Well, I'm not in a hurry. ;-) | 17:21 |
| bb|hcb | Atari-Frosch: http://dist-mirror.fem.tu-ilmenau.de/devuan this should end in /merged | 18:49 |
| Atari-Frosch | bb|hcb: thx | 18:54 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!