Hurgotron | slightly confused... I updated from beowulf to chimaera, and now wicd is gone and Network-Manager is there (as expected)... but I don't have anything in my xfce bar for the network. Do I need some applet? | 01:41 |
---|---|---|
gnarface | uh, not sure but i think if so it'd probably start with xfce in the name | 01:51 |
fsmithred | Hurgotron, check autostart apps | 01:51 |
gnarface | try "apt-cache search ^xfce" | 01:51 |
gnarface | oh, or that | 01:51 |
fsmithred | you might need to restart the desktop (or the panel) | 01:52 |
golinux | Hurgotron: Please read the chimaera Release Notes. https://files.devuan.org/devuan_chimaera/Release_notes.txt | 02:33 |
adhoc | golinux: oh, that explains much =) | 02:35 |
adhoc | golinux: did debian drop all the python2 apps from Bullseye ? | 02:36 |
adhoc | also; "Starting from Devuan 2.0 ASCII, users should exclusively use | 02:37 |
adhoc | [24]deb.devuan.org in their sources.list file, e.g.:" | 02:37 |
adhoc | that explains some weirdness too... | 02:37 |
golinux | adhoc: https://git.devuan.org/devuan?tab=&sort=recentupdate&q=amprolla | 02:49 |
golinux | There is a reason for that advice which is not weird and could save you from a reinstall . . . | 02:50 |
adhoc | I ahve been doing some work around proxying my repo access to reduce the outbound traffic | 02:52 |
adhoc | amprolla, news to me | 02:54 |
* adhoc reads | 02:54 | |
brocashelm | it's genius | 02:57 |
adhoc | ok ... | 02:58 |
adhoc | golinux: so why would any one want to use this, outside of a distro maintainer? | 02:58 |
golinux | adhoc: Exactly. "amprolla is an apt repository merger originally intended for use with the Devuan infrastructure." | 03:12 |
golinux | And it needs to be used in sources.list in order to update/upgrade a devuan installation properly. | 03:15 |
adhoc | no mention of snmpd in there | 03:22 |
adhoc | it fails to install in the post-installation step | 03:23 |
adhoc | is required for package frr-snmp | 03:23 |
adhoc | which I believe I need for the frr stack to work properly | 03:23 |
adhoc | the post-install barfs on creating a user | 03:24 |
adhoc | with a home directory of; SNMP_DIR="/var/lib/snmp" | 03:24 |
golinux | https://pkginfo.devuan.org/cgi-bin/policy-query.html?c=package&q=snmpd&x=submit | 03:24 |
golinux | So install it! | 03:25 |
adhoc | it fails on teh post install step | 03:25 |
fsmithred | looks like it installed here ok | 03:26 |
adhoc | ls -al /var/lib/snmp | 03:26 |
adhoc | total 12 | 03:26 |
adhoc | drwxr-xr-x 3 Debian-snmp Debian-snmp 4096 Apr 11 10:35 . | 03:26 |
golinux | Sorry Can't help with that. I'm the librarian . . . | 03:26 |
adhoc | so the user Debian-snmp ... | 03:26 |
adhoc | # grep snmp /etc/passwd | 03:26 |
adhoc | Debian-snmp:x:107:112::/var/lib/snmp:/bin/false | 03:26 |
adhoc | hmm | 03:27 |
adhoc | fsmithred: wondering if this is devuan specific, or a debian problem | 03:27 |
fsmithred | are you running chimaera? | 03:27 |
adhoc | yes | 03:27 |
fsmithred | me too | 03:28 |
adhoc | upgraded from 3.0 | 03:28 |
fsmithred | drwxr-xr-x 2 root root 4096 Jan 5 2019 | 03:28 |
fsmithred | I did get a warning on install... | 03:28 |
fsmithred | Warning: The home dir /nonexistent you specified can't be accessed: No such file or directory | 03:28 |
fsmithred | got that after it added group frr | 03:29 |
fsmithred | I have no user or group with snmp in the name | 03:31 |
fsmithred | I installed snmp and frr-snmp. adhoc did you do something different? | 03:32 |
fsmithred | and I installed without Recommends | 03:32 |
adhoc | decided to drop frr-snmp and will see if I can do everything I neeed to with out it for now | 03:37 |
adhoc | fsmithred: snmp would be nice, but will deal with it later =) | 03:37 |
adhoc | fsmithred: would be good to undertand why it fails to start, but it is not throwing errors when you start it on the command line, but silently fails =( | 03:42 |
fsmithred | adhoc, what are you trying to run? | 03:45 |
fsmithred | it comes with a bunch of commands | 03:45 |
fsmithred | (I don't know what any of them are or what snmp is.) | 03:45 |
adhoc | Apr 11 10:44:55 adldcns03 snmpd[9732]: Error opening specified endpoint "[::1]" | 03:45 |
adhoc | Apr 11 10:44:55 adldcns03 snmpd[9732]: Server Exiting with code 1 | 03:45 |
adhoc | ah | 03:45 |
fsmithred | where'd you get snmpd? | 03:46 |
adhoc | I wonder if this is a box I have disabled IPv6 on? | 03:46 |
fsmithred | oh | 03:46 |
adhoc | net.ipv6.conf.all.disable_ipv6 = 1 | 03:46 |
adhoc | net.ipv6.conf.eth0.disable_ipv6 = 1 | 03:46 |
fsmithred | I didn't install that package. Just snmp | 03:46 |
adhoc | from /etc/sysctl.conf | 03:47 |
adhoc | some things are so fragile | 03:47 |
fsmithred | adduser: Warning: The home directory `/var/lib/snmp' does not belong to the user you are currently creating. | 03:47 |
fsmithred | I got that warning installing snmpd | 03:47 |
adhoc | fsmithred: yes, I get that too, but it then installs anyway | 03:47 |
fsmithred | snmpd is running. | 03:48 |
adhoc | now I remember, I had to enable IPv6, reboot, install ... | 03:48 |
adhoc | some services are so fragile | 03:49 |
adhoc | because everyone runs IPv6 | 03:49 |
adhoc | even though lots of ipv6 traffic causes shitty appliances on the network to crash =/ | 03:50 |
adhoc | fsmithred: mystery solved | 03:51 |
adhoc | fsmithred: thanks for your help | 03:51 |
fsmithred | cool | 03:51 |
fsmithred | yw | 03:51 |
adhoc | next problem; get kubernetes running smoothly | 03:51 |
fsmithred | so now that I have it installed and running, what can I do with it. (in 25 words or less, please) | 03:51 |
adhoc | oh, you can query the snmp service to get status information, collect that in monitoring apps for service history | 03:52 |
adhoc | for example, on switches, NIC stats, throughput in each direction, packet errors, etc | 03:53 |
adhoc | useful when doing capacity planning | 03:53 |
fsmithred | thanks | 03:53 |
adhoc | especially when you need to prepare a report for management, "we need to upgrade" because "interfaces running at 90+% across the working day" and "staff complaining network is 'slow'" | 03:54 |
adhoc | etc | 03:54 |
adhoc | simple network monitoring protocol | 03:54 |
adhoc | now for my next error to fix for exim; 2022-03-28 13:55:49 socket bind() to port 25 for address ::1 failed: Cannot assign requested address: daemon abandoned | 03:55 |
adhoc | I see what I did there ... >.> | 03:55 |
ddsys | what folder do i create a service in? | 04:15 |
fsmithred | scripts are in /etc/init.d | 04:15 |
ddsys | thx | 04:16 |
ddsys | fsmithred: what is the command to start the service | 04:25 |
Xenguy | Something like: service <service> start | 05:02 |
Xenguy | Oh he's gone | 05:03 |
* Xenguy shrugs | 05:03 | |
adhoc | fsmithred: so if you enable IPv6 and install frr-snmp and snmpd, it just works (TM) | 05:11 |
adhoc | =/ | 05:11 |
adhoc | now to move other network appliances off to another subnet | 05:13 |
Kingsy | sorry I missed a message in here. is there logging? | 14:40 |
humpelstilzchen[ | The link in the topic might have a log | 14:53 |
ddsys | cant seem to start a service. always says down. can some help me create service? | 16:31 |
ddsys | runit | 16:31 |
fsmithred | ddsys, what service? | 16:49 |
ddsys | it can be any service. i just need to know where to create it and start it | 16:50 |
fsmithred | I have to check notes. One minute. | 16:50 |
ddsys | thanks | 16:51 |
fsmithred | you have some runscripts in a directory named after the service? | 16:51 |
fsmithred | copy that dir to /etc/sv/ | 16:52 |
ddsys | i have runscript in /etc/init.d | 16:52 |
fsmithred | if the service is using a sysvinit script, stop it. | 16:52 |
fsmithred | then run 'update-service --add /etc/sv/<service-name> | 16:53 |
fsmithred | /etc/init.d is where the sysvinit scripts are | 16:53 |
ddsys | so i dont want /etc/initd if im using runit? | 16:54 |
fsmithred | yeah, you do | 16:54 |
ddsys | oh | 16:54 |
fsmithred | in debian/devuan runit uses sysvinit scripts by default | 16:54 |
ddsys | ok | 16:54 |
fsmithred | except for the gettys and a few other things | 16:54 |
fsmithred | but you can switch to runit scripts | 16:55 |
fsmithred | https://dev1galaxy.org/viewtopic.php?id=3716 | 16:55 |
fsmithred | and there are sources for scripts from the runit maintainer, from antix and a couple other places. | 16:55 |
ddsys | fsmithred: thanks for your help i will try your update-service command | 16:56 |
fsmithred | look through the discussion I posted | 16:56 |
fsmithred | also be aware that void linux and some other sources of info might say to put things in different directories. Modify those instructions to fit the debian/devuan locations. | 16:58 |
ddsys | fsmithred: yes i noticed different sources says to use diff directories | 17:05 |
ddsys | i was able to add service but still down | 17:06 |
fsmithred | ddsys, what service, and can you paste your runscript at paste.debian.net? | 17:20 |
ddsys | paste.debian.net/1237600 | 17:23 |
ddsys | it prevents staircase effect on shutdown | 17:24 |
ddsys | i had the script working but i want it as a service so i can learn | 17:24 |
ddsys | i put script in /etc/sv/staircase/run | 17:28 |
ddsys | then update-service --add /etc/sv/staircase | 17:28 |
ddsys | then sv up staircase and then sv status staircase but still says down | 17:29 |
fsmithred | what is the script supposed to do? | 17:30 |
fsmithred | and why sudo? system scripts are run as root | 17:31 |
ddsys | have you ever heard of staircase effect? | 17:31 |
ddsys | oh ok about sudo | 17:31 |
fsmithred | no | 17:31 |
ddsys | on shutdown the messages are all zigzagged | 17:32 |
ddsys | instead of in a straight column | 17:32 |
ddsys | most people dont care but im trying to learn | 17:33 |
fsmithred | I'm not the best person to be helping with managing services, but that seems like something that doesn't need to be running all the time. Just at shutdown. | 17:34 |
ddsys | lightdm causes it | 17:34 |
systemdlete | fsmithred: I made a typo at the end of the refracta 11 refractainstall script at the point where it asks which device (not devices) to install the bootloader (which I thought I had already answered the question before, idk). Anyway, I was able to reboot but the resolution is awful. Are the radeon drivers on the refracta ISO also? | 17:34 |
ddsys | needs at startup foe some reason | 17:35 |
fsmithred | systemdlete, firmware-amd-graphics should be in the firmware dir | 17:35 |
systemdlete | Is that the only package I will need? | 17:35 |
fsmithred | I hope so | 17:35 |
systemdlete | ok, thanks | 17:36 |
fsmithred | no network? | 17:36 |
fsmithred | ddsys, I don't know how to help, but if you poke around the forum and look at the discussions about runit, you will learn a lot. The debian runit maintainer sometimes posts on those threads. | 17:37 |
ddsys | ok thx | 17:37 |
systemdlete | fsmithred: I still think the installer should be slightly more robust, and less confusing with the grub steps. | 17:38 |
fsmithred | no shit | 17:38 |
systemdlete | I'm not trying to give you shit, honest. | 17:38 |
systemdlete | Just some (hopefully) constructive feedback. | 17:38 |
fsmithred | yeah, I am agreeing with you. | 17:38 |
systemdlete | I know. | 17:38 |
fsmithred | I think I changed the button label in the last version from 'copy files' to 'install bootloader' | 17:39 |
systemdlete | other than that, the install was pretty quick and easy. | 17:39 |
systemdlete | "button" -- I didn't see many buttons, just numbered prompts. | 17:39 |
systemdlete | (did I invoke the installer incorrectly?) | 17:40 |
systemdlete | Also, I note that the installer did not seem to be available on the desktop. | 17:40 |
systemdlete | I thought it was the last refracta release | 17:40 |
systemdlete | But using my brain, I figured out to invoke "refractainstaller" right from the cmd line | 17:40 |
fsmithred | refractainstaller gives you the cli script, refractainstaller-yad gives you the graphical version. | 17:45 |
fsmithred | it's in the menu under System | 17:45 |
Atari-Frosch | TIL: After reboot of my server that – besides other services – runs a Jitsi instance, Jitsi didn't accept any connections. Reason: prosody. With the reboot it was started using service, not prosodyctl. Killed the process (stopping by service didn't work), restarted with prosodyctl, my Jitsi is back. | 18:57 |
Atari-Frosch | Server runs with Devuan Beowulf. | 18:59 |
Atari-Frosch | (Yeah, I know, I should update.) | 18:59 |
brocashelm | if someone is on ceres and decided to switch back to stable, would it make sense to switch to daedalus (testing) and tough it out until it's stable? so as to not have to go through reinstalling and restoring/fixing configs all over again | 22:51 |
gnarface | downgrading isn't a supported action by upstream | 22:53 |
gnarface | i tried it once and it appeared to work at first but subsequent updates tore the install apart | 22:54 |
gnarface | ymmv but i'd advise doing a backup and reinstall; it'll actually be faster by a lot | 22:54 |
gnarface | lots of configs aren't gonna be compatible anymore either with older versions | 22:55 |
gnarface | so even if your package dependency tree isn't corrupted you're still probably in for approximately the same amount of manual config restoration | 22:56 |
gnarface | however, at some point recently i probably would have said daedalus and ceres are still close enough that you might succeed in that plan - i don't know if that's the case still | 22:57 |
gnarface | there was a point shortly after daedalus was first cut that they were almost identical | 22:58 |
gnarface | maybe someone else around here would have a better idea if that's the case but a minimal install only takes half an hour or so | 22:58 |
rwp | Going from Unstable to Stable is basically everything gnarface said. Going from Unstable to Testing probably is not as bad. | 23:00 |
rwp | I suggest installing apt-show-versions, configuring both ceres and Testing, and seeing how different they are from each other using apt-show-versions. | 23:00 |
gnarface | hmm, interesting plan | 23:00 |
fsmithred | brocashelm, I've done it, and if you time it right, it's no problem. | 23:00 |
fsmithred | since bookworm has not frozen yet, you will be safe | 23:01 |
fsmithred | just change sources to daedalus, and you will get the same updates you've been getting, only a week later. | 23:01 |
fsmithred | going to chimaera would be a major pain in the ass. You have to play with pin priorities (1001 on chimaera) | 23:02 |
fsmithred | and cross your fingers | 23:02 |
fsmithred | and have some installation media on hand for when you're done playing | 23:02 |
rwp | There are two categories for differences. The 10 day waiting period for migration from Unstable to Testing. Packages stuck from migration due to RC bugs logged against them preventing migration but where Unstable keeps moving. | 23:03 |
fsmithred | you could keep a line for ceres and pin it low so that you only get stuff from ceres if you ask for it. | 23:05 |
rwp | I think any downgrade that includes glibc, which would almost certainly include a long list of other libs too, would be REALLY HARD. Technically possible. But much easier and much more reliable to reinstall. | 23:05 |
fsmithred | he doesn't need to downgrade at this point | 23:06 |
fsmithred | just wait | 23:06 |
rwp | ":51 <brocashelm> if someone is on ceres and decided to switch back to stable..." Sounded to me like Stable was the target. :-( | 23:06 |
fsmithred | yeah, that was his first thought. | 23:06 |
fsmithred | that would be bad. | 23:06 |
fsmithred | but his current ceres will soon be daedalus | 23:07 |
rwp | I totally agree that moving from Ceres to Testing would be very easy, no problem, no muss, no fuss. | 23:07 |
fsmithred | brocashelm, what's the issue? Is something broken, or are you just tired of all the updates? | 23:07 |
brocashelm | ceres works fine for me and i'm fine with updating. i was just curious about turning one of my machines into daedalus instead so that it'd soon turn to stable | 23:36 |
brocashelm | only packages i have held back right now are for elogind (four of them), but that will be fixed soon | 23:37 |
brocashelm | and, yeah, daedalus is still close to ceres | 23:38 |
brocashelm | i just figured it'd be a matter of giving it time for packages to switch to their daedalus counterparts once there are newer versions than the ones you got from ceres | 23:38 |
rwp | brocashelm, It's really no problem at all for your migration plan from Ceres to Daedalus. Easy. No worries at all. | 23:43 |
brocashelm | nice. i figured that would be simple enough. i just wanted to be sure, in case i plan on going back to stable (again, having zero problems with ceres in the two years i've been running it as my daily driver, and i update all the time) | 23:48 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!