22[00:15:09] <Casper26> Anyone seen an issue with gnome-connection-manager 1.2.1 when lauched and a console is selected it never connects just has a blinking cursor?
62[01:20:27] *** Quits: fax (~quassel@replaced-ip) (Remote host closed the connection)
63[01:20:33] <Lady_Aleena> !bullseye
64[01:20:33] <dpkg> The release following Debian 10 "Buster" is codenamed "Bullseye" (Woody's horse in Toy Story 2) and will be Debian 11. It is the current "testing" release. Remember that straight after a stable release, all sorts of mess suddenly lands in "testing" and it is best avoided if you don't like debugging things. replaced-url
65[01:21:52] <Lady_Aleena> Looks like the link is outdated.
74[01:28:34] <Lady_Aleena> themill, I went back a year and didn't find anything. *shrugs* Oh well. Maybe just take out the link to the freeze information for Stretch?
75[01:29:01] *** debhelper sets mode: +l 1125
76[01:31:53] <Lady_Aleena> I'm just eager to see a little thing I suggested added to a command in util-linux in action, but I don't think it will get to my system until bullseye.
83[01:38:18] <themill> it would be very unusual for that to appear earlier than bullseye
84[01:38:28] <Lady_Aleena> But it might not make it into bullseye because it was added only a little over a month ago.
85[01:40:26] <themill> That depends partly on whether its developers intend it for 2.37 or 2.36.2. (There's a reasonable chance of 2.37 hitting bullseye, looking at their typical release pattern)
86[01:42:01] <phogg> hard to be sure until the upstream release is out
87[01:42:10] <Lady_Aleena> I just feel a little tickled my small request will be part of the something people will use.
113[01:58:22] <themill> There's basically zero info out there -- the d-d-a announcement and the wiki will do
114[01:59:35] <Lady_Aleena> themill, you know more than I do. 8)
115[02:00:44] <themill> updated
116[02:01:02] <Lady_Aleena> themill, thank you very much!
117[02:01:09] <Lady_Aleena> !bullseye
118[02:01:10] <dpkg> The release following Debian 10 "Buster" is codenamed "Bullseye" (Woody's horse in Toy Story 2) and will be Debian 11. It is the current "testing" release. Remember that testing is called testing for a reason; good bug reports with patches are greatly appreciated! replaced-url
149[02:44:33] <s_> is there a debian+1 channel or similar
150[02:44:43] <sney> !debian-next
151[02:44:44] <dpkg> #debian-next is the channel for testing/unstable support on the OFTC network (irc.oftc.net), *not* on freenode. If you get "Cannot join #debian-next (Channel is invite only)." it means you did not read it's on irc.oftc.net. See also replaced-url
152[02:44:50] *** Quits: victorpr (~victorpr@replaced-ip) (Remote host closed the connection)
153[02:47:01] * EdePopede waiting
154[02:47:46] * dvs is getting ready to complete the BASIC FOR loop...
168[03:03:39] <sney> debian was intending to remove python2 but some popular software refused to switch. so instead it was moved into explicit python2-* packages. python3 is still the new default.
176[03:09:29] <spacedust> im using debian in a vm, and when i send an acpi shutdown it does nothing, i should like it to shutdown when i send it a shutdown message from virt-manager
177[03:09:38] <spacedust> how can accomplish that ?
219[03:41:10] <spacedust> Specifying boot ID or boot offset has no effect, no persistent journal was found.
220[03:41:13] <spacedust> thats what i get
221[03:42:01] <jmcnaught> spacedust: /usr/share/doc/systemd/README.Debian shows how to make a persistent journal. Basically you just create /var/log/journal
222[03:42:06] <spacedust> is there a way to tell libvirt it should wait until it shuts down ? or it does wait ... or until the wait time is off ?
229[03:47:13] <spacedust> considering the host reboots fairly fast even with the added time to wait for vm shutdown ... i think the vm shuts down then tells that to the host then that reboots
230[03:47:27] <spacedust> i changed wait time to 500
256[04:04:58] *** Quits: metro (~metro@replaced-ip) (Quit: Leaving)
257[04:05:28] <jmcnaught> SHUTDOWN_TIMEOUT controls how long to wait for guests to shutdown, but it should not wait the whole default 300 seconds if all guests are shut down.
396[07:39:06] <Lope> so my upgrade from buster+backports to bullseye yesterday failed miserably.
397[07:39:38] <zerocool> can you do a bond-mode 2 of two bond-mode 4's?
398[07:40:11] <zerocool> like active/backup of two LACP
399[07:40:18] <Lope> ratrace, as you predicted there were issues with reverting the backup over my live system. So it didn't boot. But I just connected the SSD to another system and then restored the backup with rsync. So I'm back to buster+backports
400[07:41:36] *** Quits: Deknos (~deknos@replaced-ip) (Remote host closed the connection)
401[07:42:41] <zerocool> oh shit yeah, you can next bonded interfaces exactly as described.
404[07:44:28] <Lope> a guy at #debian-next reckons you can't upgrade buster+backports to bullseye.
405[07:45:24] <Lope> in terms of backports, I've got nvidia, realtek network DKMS, ZFS, and kernels.
406[07:46:29] <Lope> In theory I should be able to boot an old kernel, then remove all of the above, and backports from sources, then reinstall the above, and I should be back to a pure buster?
407[07:46:33] <ryouma> what is the normal thing to do with backports and backports sources?
408[07:46:38] <ryouma> when dist-upgrading?
409[07:46:40] <Lope> which should be upgradable to bullseye?
411[07:47:26] <Lope> ryouma, some guys here yesterday said it should upgrade fine, but my system only managed to do the upgrade, but failed to do the dist-upgrade
412[07:47:32] <Lope> due to dependency conflicts
413[07:47:43] <Lope> Something wanting an old version and something else wanting a new version.
414[07:47:54] <ryouma> ah and your backports might have been one of those
415[07:48:07] <Lope> yeah, that makes sense in theory.
416[07:48:54] <Lope> anyway, when I'm in the mood for trouble I'll attempt to go back to stock, then attempt the upgrade again.
417[07:49:15] <Lope> Maybe I'll even use a USB ethernet adapter so I don't need the realtek DKMS module
418[07:49:21] <Lope> (during the upgrade)
419[07:49:52] <Lope> perhaps I can do the upgrade without nvidia, realtek DKMS or ZFS installed.
420[07:50:59] <Lope> This failed upgrade experience has inspired me to try btrfs instead of ext4 for my hypervisor's rootfs
421[07:51:11] <Lope> because I'd be able to easily snapshot
422[07:52:34] <Lope> Pretty cool to see that btrfs can be shrunk, (like ext4)
423[07:53:00] <Lope> the ability to shrink is nice to have for a rootfs
424[07:53:12] <Lope> (in contrast to XFS, which can only expand)
433[07:59:43] *** Quits: yuta (~pi@replaced-ip) (Quit: WeeChat 2.9)
434[07:59:45] <Lope> Well, I did "snapshot" with rsync, and it allowed me to un-break my system
435[08:00:25] <Lope> ryouma, But it's still more of a hassle and less efficient to snapshot with ext4 and rsync than what btrfs sounds like.
436[08:01:41] <Lope> To snapshot I made a new fs on my zfs, then rsync'd my live system into that, then created a snapshot of that zfs to protect it from a casual mistaken change.
437[08:02:01] <Lope> But ZFS is a bit iffy on linux. Not great for backing up your rootfs.
438[08:02:15] <Lope> if anything goes wrong with DKMS then you've got no access to ZFS
439[08:02:28] <Lope> and need to connect your stuff to an external system that ZFS is working on.
440[08:03:10] <Lope> I mean I could obviously have rsync'd to anything, but still, it doesn't really make sense to use ext4 if you're going to be snapshotting.
444[08:07:56] <Lope> I know some people use ZFS for the rootfs. I've never done it. I'll probably try it at some point. It does seem like a good idea since I already use ZFS everywhere I can anyway.
445[08:10:13] <ryouma> idk what dkms is or zfs's issues (i got annoyed with it for my purposes) and i share the interest in btrfs, but have heard mixed reports on the reliability of btrfs, is all.
447[08:10:20] *** Quits: ivz_hh (~ivzhh@replaced-ip) (Remote host closed the connection)
448[08:11:11] <ryouma> maybe it's ok these days, dunno
449[08:11:40] <Lope> I heard horror stories about btrfs when I first got into linux like 8 years ago, but from what I can tell btrfs has come a very long way and is in widespread use.
450[08:11:55] <ryouma> oh kernel modules
451[08:11:57] <Lope> I think as long as you don't use raid5 or raid6 modes you should be rock solid.
452[08:12:09] <Lope> Just don't do anything too weird and unusual.
453[08:12:31] <Lope> If I did anything special with it, it would just be mirroring and snapshotting.
472[08:18:59] <ryouma> (not useful for your current issue but that would solve the fact that currently rysnc will sometimes not copy over a file that did change, because something sets mtime to the same time (a couple of debian packages do this in their files for some reason))
480[08:26:53] <b4d> jelly: 10.6 with backports, now cyrus-imapd is on version 3.2.3-1~bpo10+1 and won't start
481[08:31:27] *** Quits: flakE (~flakE@replaced-ip) (Quit: Every once in a while I like to stick my head out of the window and smile for a satellite picture.)
486[08:41:39] <Lope> just trying to understand the requirements here: replaced-url
487[08:41:45] <Lope> Building binary packages for a single kernel flavour requires up to 15 GB space in the package directory and 300 MB in /tmp (or $TMPDIR).
488[08:41:54] <Lope> Building all binary packages for i386 or amd64 currently requires about 50 GB space in the package directory.
506[08:49:02] <Lope> okay, I've read the page again, I think when it says "packages" it means "kernel packages" because that is in headings above. I think "kernel packages" means the .deb kernel packages that one sees in `dpkg --list | grep linux-image` `apt-cache search ^linux-image` etc. So when they say all packages, they mean compiling every debian kernel that's available for your current release.
519[09:10:02] <b4d> jelly: thanks for the nudge in the right direction, I have added buster-backports point in time repo from snapshots and downgraded the affected packages, works now
531[09:14:37] <Lope> I need to downgrade my buster-backports stuff to buster... I was planning to just boot the old kernel, then remove --purge the backports packages, remove the backports source, update and hope the system is then fully stock.
532[09:14:40] <jelly> Lope, they picked a slightly older backport from snapshot.d.o
555[09:37:32] <ratrace> Lope: you don't say.... :)
556[09:38:00] <ratrace> Lope: you left before I could explain the problem of reverting the files under running processes, esp. if the files belong to previous versions of the processes.
559[09:38:54] <Lope> ratrace, well, now that I destroyed my system first hand, I'll most likely remember the perils of overwriting a live running system with a backup.
560[09:40:04] <jelly> Lope, in general no, downgrading is not supported
561[09:40:47] <jelly> you can apt install foo=version bar=version2, but going to a lower version is never tested
562[09:41:35] <jelly> so there is apt / aptitude syntax to do it, if that's what you asked
563[09:41:56] <ratrace> the primary problem is with postint scripts that do changes that apt has no actions to undo. otherwise, a "downgrade", in its purest form, is like "upgrade", just a replacement of files. if there are steps in upgrade that require more than just that, the downgrade will probably break.
564[09:42:14] <Lope> jelly, do you recomment install foo=version OR apt remove --purge foo && apt install foo ?
565[09:44:41] <Lope> jelly, ratrace: are you familiar with the fact that ClearLinux doesn't run on CPU's older than Sandybridge and performs better than other distros? I'm wondering if I can compile the debian kernel with the same optimizations. Do you know specifically which options would relate to "only sandybridge and newer" replaced-url
568[09:45:14] <Lope> Because when I've compiled kernels in the past, I've never seen any options for selecting "sandybridge and newer". Only stuff like supporting atom and via crap etc.
569[09:45:18] <jelly> Lope, I recommend exercising caution. Usually a mail server has important configuration, can you really purge it without thinking?
576[09:49:16] <ratrace> Lope: I don't think the kernel options have much to do with cpu gen, other than enabling hardware support for things like crypto, crc, and similar
577[09:50:01] <ratrace> Lope: otherwise yes, I know what CL is, but their performance enhancements come through compiler flags and patches to the compiler and other source that maximize usage of intel specific behaviour
588[09:58:27] <Lope> ratrace, oh, interesting, that concurs with my understanding.
589[09:59:02] <Lope> ratrace, because I figured the "cpu support, etc" options in the kernel build menu were relating to what code to compile, rather than how to optimize code.
590[09:59:44] <Lope> so if you add support for a crappy old CPU, then it just adds extra code for the old CPUs, doesn't make stuff slower when running on newer CPUs.
591[10:00:26] <ratrace> Lope: perhaps you're talking about some options specific to the CL kernel, but that relate to compiler flags? Eg. gentoo has "experimental" patches that add on choosing cpu generation or "march=native" but that really alters the compiler options
593[10:01:14] <ratrace> Lope: depends what kind of "support" it is. in most cases it'll be the opposite of what you said. instruction set optimization specific to older CPUs that do not have faster/better functionality of newer CPUs
594[10:01:19] <Lope> ratrace, can you see what/where the compiler optimizations are? replaced-url
595[10:02:14] <ratrace> thoes are patches, not CONFIG_ options, I thought you talking about those.
596[10:02:23] <Lope> ratrace, is march=native the best you can do?
597[10:02:34] <Lope> in terms of optimizations
598[10:02:50] <ratrace> yes
599[10:03:03] <ratrace> but then that code is locked to that CPU (gen and model) in most cases
601[10:03:18] <Lope> ratrace, well, I'm open to all possibilities or how the debian kernel can be compiled to run faster on a particular system.
602[10:04:10] <Lope> ratrace, well, I don't see a problem with compiling a unique kernel for each of my computers. it only takes a few hrs and is automated. that means when I use the computer, it'll be faster. So that's worthwhile.
603[10:04:22] <ratrace> Lope: unfortunately there is no single solution. even if you rebult all with march=native, that in itself is only going to marginally help, maybe a few %
605[10:05:18] <ratrace> Lope: in addition you can enable various hardware supported features of programs via their build-time config, and this mostly applies to multimedia apps that can take advantage of special instruction sets, and otherwise don't autodetect at run time
614[10:13:35] *** Quits: nobyk (~nobyk@replaced-ip) (Remote host closed the connection)
615[10:14:04] <ratrace> Lope: yes you can specifcy the march= and mtune= as well, but if you're compiling on that CPU itself, then march=native is better
616[10:14:51] <Lope> ratrace, thanks. Can you specity -march when you run the `make nconfig` command? replaced-url
617[10:15:33] <Lope> I've not compiled any debian kernels yet. I compiled a lot of ubuntu kernels a few years ago (had to edit source to fix a driver problem)
618[10:16:19] <Lope> so `make nconfig -march=native`
619[10:16:25] <ratrace> Lope: it's a CFLAG, so I'm not quite sure how to pass that to debian kernel building options
620[10:16:50] <Lope> Previously I also used to do -j 4 for quadcore. But I've heard that the default is to use all threads now, making -j unnecessary.
621[10:17:07] <ratrace> I don't customize Debian like that, I use Debian Stable Vanilla because Stable. For everything else I use Gentoo where all this "native" to the distro and is easy to manage
630[10:18:51] <Lope> thanks for the invite though :)
631[10:19:58] <Lope> ratrace, you know the dropbear initramfs luks thing... can gentoo do all that?
632[10:19:59] <ratrace> Lope: "best tool for the job" and all that. if you really want to take advantage of modifying build time options for packages, then gentoo is really your best tool for that.
638[10:22:17] <radkos> Hello. I 'm receiving the following error while botting up the system and I have no network prior to executing manually service networking start
639[10:22:18] <radkos> [ 5.105831] systemd[1]: network-online.target: Job networking.service/start deleted to break ordering cycle starting with network-online.target/start
640[10:22:27] <ratrace> Lope: patches included, you just drop them into designated dir in /etc, and teh package manager slurps them in when installing a package :))
641[10:23:15] <ratrace> radkos: so what change to service dependencies did you make?
643[10:23:49] <Lope> ratrace, do you mean in gentoo you can specify -march=native in your /etc/ and then any package you install gets built with that optimization?
644[10:24:16] <Lope> ratrace, how reliably do packages build in gentoo?
645[10:24:39] <Lope> ratrace, when I've had to manually build stuff in linux, it only works 95% of the time.
646[10:24:49] <ratrace> Lope: no, you add CFLAGS="-march=native" to your /etc/portage/make.conf. but I mean patches. you can add patches to software by dropping them into a designated dir and the package managers takes them in when building it. really, pop into #gentoo or #gentoo-chat, for more info :)
647[10:24:52] *** Quits: udet (~RalphBa@replaced-ip) (Remote host closed the connection)
648[10:24:55] <Lope> Which could be annoying if you're installing 200 packages.
649[10:25:06] <radkos> I am just having a glusterfs mount entry in /etc/fstab nothing directly changed in service conf
653[10:25:48] <Lope> ratrace, cool :) okay will do when I'm ready for a bigger beard :)
654[10:25:50] <ratrace> Lope: that's the problem with building on debian. when an update comes you have to repeat all of it over again, and you don't even know there's an update because you built from src deb
753[11:09:36] <absinthe> hello. i've been asked to upgrade a jessie server to buster, but the main issue i'm facing is that this server seems to already mix jessie/stretch repos; so im not really sure how to proceed.
754[11:09:48] <ksk> dob1: I would expect it to work, yes. I dont know for sure though, sorry.
755[11:10:02] <ksk> Maybe you can google which min kernel verison Intel XY does need, was not really able to do so.
756[11:10:19] <dob1> ksk, ok but on backports what is the name of linux kernel package? I am search linux-image but I don't find it
758[11:10:34] <absinthe> Id usually just upgrade to latest $version then dist-upgrade sequencially, but here... it seems i should directly jump to stretch in order to put stuff on the same level ?
766[11:13:01] *** Quits: leden (~leden@replaced-ip) (Remote host closed the connection)
767[11:13:11] <ksk> absinthe: you are basicly on your own here, as you have created
768[11:13:13] <ksk> !frankendebian
769[11:13:13] <dpkg> When you get random packages from random repositories, mix multiple releases of Debian, or mix Debian and derived distributions, you have a mess. There's no way anyone can support this "distribution of Frankenstein" and #debian certainly doesn't want to even try. Ask me about <reinstall>
770[11:13:39] <ksk> What you say sounds reasonable, I dont have any idea if it will work though. Good Luck ;)
775[11:19:40] <ratrace> radkos: I see. I thought maybe you had something else so you entered a circular dep of network needing the part of fs that requires network. try using x-systemd.after=network-online.target instead of _netdev .
790[11:30:20] <wrksx_> Hey. I'm experimenting with OnVif discovery, which is a protocol for device detection on a network, and I've got firewall issues. The discovery starts with sending a UDP 'probe' on a specific multicast address, and responses are expected on the interface and port that sent the probe. I believe the firewall is blocking responses since they are coming from the IP of the devices, and not from the multicast address, which is not
791[11:30:21] <wrksx_> playing nicely with the 'statefulness' of the firewall. Do you have any idea how to solve that?
800[11:39:47] *** Quits: BenNZ (~Ben__@replaced-ip) (Quit: Everytime I think IQ's must have dropped recently, I remember that this is the internet)
801[11:40:42] <radkos> ratrace, I've tried with x-systemd. options not long ago on the same system and the networking didnt start. But I will give it another try.
822[12:02:32] <Lope> ratrace, audio working with normal 5.8 backports kernel
823[12:03:03] <Lope> hey guys is rsync -aAXS enough to backup some program's files perfectly?
824[12:03:09] <ratrace> Lope: kinda makes sense, those kernels are designed for virtualized server workload where sounds is likely never gonna be needed
825[12:03:23] <Lope> ratrace, for sure :)
826[12:04:00] <ratrace> Lope: yes, -aAX sounds sane enough. dunno about -S, I wouldn't.
827[12:05:02] <Lope> I've got a script that sets up a tmpfs, rsyncs data from a SSD dir to the tmpfs dir, uses it, then rsyncs it back to the dir on disk afterwards.
828[12:05:26] <Lope> Lately somethings been going wrong where it's not backing up a lot of the data. I wonder if it's symlink related.
873[13:02:06] <shmoobert> is it possible to do a checksum of a /dev/? I maade an image of a dev and then tried sudo sha256sum /dev/nvmeXX and it tells me permission denied? shouldnt sudo take care of that?
874[13:02:24] *** Quits: de-facto (~de-facto@replaced-ip) (Remote host closed the connection)
949[14:04:01] <shtrb> Is there some ineherted problem to use at the same time an rfcomm connection for ip communitcatoin over bluetooth and passing audio via HPF(?) to pulseaudio ? If connect with rfcomm pulseaudio no longer play audio from my device.
954[14:06:39] <dpkg> A backport is a package from a newer Debian branch, compiled from source for an older branch to avoid dependency and <ABI> complications. replaced-url
1019[14:46:28] <ozgur> ratrace, if i turn off UEFI USB will boot legacy but this time UEFI will not boot do i have a chance to change UEFI after non-UEFI?
1020[14:46:32] <ratrace> dob1: yes, 5.7.10 for buster
1021[14:46:52] <ratrace> ozgur: you either boot in EFI mode or legacy/bios mode
1022[14:46:55] <dob1> ratrace, sorry but how do you find it here ? replaced-url
1023[14:47:24] <n4dir> enter linux-image, hit enter to search, pick buster-backports in the list
1024[14:47:26] <ratrace> !backports
1025[14:47:26] <dpkg> A backport is a package from a newer Debian branch, compiled from source for an older branch to avoid dependency and <ABI> complications. replaced-url
1026[14:47:57] <n4dir> ups, i meant packages.debian.org.
1027[14:48:04] <ozgur> ratrace, okay when i installed legacy mode and installed Debian, can i do uefi later?
1032[14:48:45] <dob1> ok you have to search linux-image-amd64 I was searching linux-image
1033[14:48:52] <ratrace> ozgur: should be possible, but when you install debian don't forget to reserve/leave a partition that will later become ESP FAT32
1034[14:49:03] <dob1> but the "Only show exact matches: " makes no sense then
1035[14:49:21] <ozgur> ratrace, hm okay
1036[14:49:23] <ozgur> thanks
1037[14:49:46] <n4dir> dob1: i see. Didn't know that problem exists
1069[15:28:31] *** Quits: swine_ (~swine-___@replaced-ip) (Remote host closed the connection)
1070[15:28:32] <shmoobert> hmmmm, re-ask: is it possible to do a checksum of a /dev/? I maade an image of a dev and then tried sudo sha256sum /dev/nvmeXX and it tells me permission denied? shouldnt sudo take care of that?
1099[15:38:40] <greycat> The obfuscated command is a giant red flag for me. What else was not reproduced correctly?
1100[15:38:49] <greycat> Why would you have to obfuscate the device node?
1101[15:39:01] <rk4> my nvme enumeration is classified!
1102[15:39:28] <fling> what are you guys doing?
1103[15:39:48] <rk4> we're not really sure.
1104[15:40:12] <fling> I suggest to stop running commands on your devices ;>
1105[15:40:17] <fling> or you end up as linus
1106[15:40:53] <rk4> but how will i collect my UUoC award
1107[15:42:38] <shmoobert> I think its working? will take awhile i guess
1108[15:42:57] <rk4> so what was the permission defined |:
1109[15:43:19] <greycat> I'm assuming we were not given the actual failing command.
1110[15:43:46] <greycat> The failing command was probably sudo sha256sum < /dev/something but they changed it to "sudo sha256sum /dev/something" in IRC
1111[15:44:13] <ratrace> teh plot thickens!
1112[15:44:20] <greycat> All of this wasted time could be avoided if they'd just paste the actual failing command.
1113[15:45:15] <shmoobert> rk4, I think i was specifying the wrong dev :/ I was thinking it would be like nvme0 but when i did df it seems that i should have been doing nvme0n1 and it seems to be working
1114[15:45:38] <rk4> yes but one day the episode of tech support detective will have an interesting plot twist, you'll see!
1115[15:46:03] *** Oksana_ is now known as Oksana
1116[15:46:19] <shmoobert> regardless, help was appreciated
1132[15:57:40] <JordiGH> I'm pretty sure I'm not on an alsa-only system, but Google Meet definitely had no sound for me.
1133[15:57:54] <ksk> !debian-next
1134[15:57:54] <dpkg> #debian-next is the channel for testing/unstable support on the OFTC network (irc.oftc.net), *not* on freenode. If you get "Cannot join #debian-next (Channel is invite only)." it means you did not read it's on irc.oftc.net. See also replaced-url
1148[16:05:32] <Onyx47> greycat: I have seen a FoxPro database being managed in Excel through OLE or whatever it's called, and that was less than 5yrs ago... honestly, I'd rather take Access at that point (or, you know, go get yourself a proper DB...)
1149[16:06:25] <shtrb> Give them grep , sed , and vim , let them edit text files.
1150[16:06:57] <shtrb> or create a systemd wrapper , and send that as logs for some unit file. like nhs.unit
1168[16:25:48] *** Quits: ohnx (~ohnx@replaced-ip) (Quit: No Ping reply in 180 seconds.)
1169[16:26:14] <jak2020> i want to copy a file from server 1 to server 2 via commands, and every day at 6am start the copy...... i know the way is: keys any tutorial?
1170[16:26:51] <greycat> create an ssh key pair with no passphrase, and cross your fingers that nobody steals the private key, since it won't be protected by a passphrase
1171[16:28:51] <jelly> asterismo, why do you think there's something wrong? What's the Installed-Size?
1172[16:29:13] <jak2020> first on server 1 (original) ssh-keygen -t rsa -b 4096 -C "jak2000@gmail.com"
1173[16:29:44] <jak2020> and then ssh-copy-id copy the key from server 1 to server 2 right?
1174[16:29:49] <jelly> then ssh-copy-id server2, yes
1224[16:53:43] <Sentinel> no it's only console colors, not graphics
1225[16:53:53] <Sentinel> is there a night mode in bash?
1226[16:54:20] <greycat> bash doesn't do ANYTHING with colors except partial tab completions and even those are off by default; you mean is there a night mode in the *LINUX CONSOLE*.
1227[16:54:47] <Sentinel> yes I mean that
1228[16:54:48] <greycat> are you sure it's not a hardware issue? monitor isn't fully warmed up, or cable is loose?
1229[16:54:50] <Sentinel> what's the answer?
1230[16:54:54] <Sentinel> yes I'm sure
1231[16:55:31] <Sentinel> like... are the colors defined in a file? which package could have changed that file?
1277[17:17:39] <Haohmaru> sigint and then on the other hand, you may have been looking at a certain thing your whole life, and then someone points out a detail, and you hadn't noticed it before, and it's right there
1278[17:17:56] <Haohmaru> eyeballs+brain does mysterious things
1279[17:21:04] *** Quits: shingouz (~not@replaced-ip) (Disconnected by services)
1302[17:25:49] <jelly> Sentinel, are you perhaps using a non-free graphic driver that hates the console, like nvidia?
1303[17:25:54] <Sentinel> yeah
1304[17:26:00] <Sentinel> but it worked fine before
1305[17:26:07] <jelly> yeah, it's not consistent
1306[17:26:13] <Sentinel> I'm guessing it's a system setting though
1307[17:26:17] <Sentinel> like "enable VTs"
1308[17:26:28] <jelly> basically don't count on VT working if you use nvidia driver
1309[17:26:53] <Haohmaru> aww, crapvidia
1310[17:27:14] <jelly> sometimes they work. Sometimes the text mode isn't restored properly
1311[17:27:32] <Haohmaru> it internally tosses a coin
1312[17:27:38] <jelly> monitor switching off as if no signal suggests the latter is happening
1313[17:28:23] <Haohmaru> so the issue was actually colors in the textmode thing, not one of those GUI terminals?
1314[17:28:35] <Sentinel> no it was in a gui terminal
1315[17:28:37] <jelly> two different symptoms
1316[17:28:43] <Haohmaru> hm
1317[17:28:44] <Sentinel> sorry I was unclear earlier
1318[17:29:03] <Sentinel> I confirmed it was a gnome-terminal problem by trying xterm
1319[17:29:11] <Sentinel> but now I can't see other VTs
1320[17:29:13] <jelly> Sentinel, that's why we asked whether you were actually talking about the console, or about a terminal emulator app
1321[17:29:19] <Sentinel> yeah sorry
1322[17:29:24] <Haohmaru> coz if your default terminal was changed (this can happen as you install stuff or change settings) you may run a different terminal which has different settings.. etc..
1323[17:29:44] <jelly> "can't see VT" is somewhat to be expected with nvidia, if you have to use it, live with it
1324[17:29:47] <greycat> Haohmaru: they were pretty adamant that they're not using X or Wayland. Pure raw console.
1325[17:30:01] <Sentinel> I don't like it :(
1326[17:30:14] <Haohmaru> greycat i didn't read carefully ;P~
1330[17:31:14] <greycat> jelly: it... was all a lie?
1331[17:31:24] <Haohmaru> >:)
1332[17:31:28] *** Quits: frgo (~frgo@replaced-ip) (Remote host closed the connection)
1333[17:33:18] <jelly> Sentinel, I _think_ nvidia actually uses kms now to set up a framebuffer and memory management on the kernel side, but they do not provide a fb console driver, just X
1343[17:36:25] <dka-> I want to upgrade jessie to buster
1344[17:36:32] <Sentinel> jelly: yep found them
1345[17:36:33] <dka-> I am reviewing the installed backport plugins
1346[17:36:34] <greycat> gotta do stretch first
1347[17:36:41] <dka-> Yes of course
1348[17:36:48] <jelly> dka-, no skipping releases, you have to jessie->stretch frist
1349[17:36:49] <dka-> I need to remove a few things before trying to dist upgrade
1350[17:36:50] <Sentinel> jelly: I'm currently trying to see about this replaced-url
1351[17:36:59] <Sentinel> the "NAutoVTs" setting
1352[17:37:09] <dka-> jelly, OK, so if I remove all bp installed packages. how can I upgrade to stretch?
1353[17:37:21] <dka-> I assume it would be: apt-get update apt-get upgrade apt-get dist-upgrade
1354[17:37:23] <jelly> Sentinel, that doesn't really matter if nvidia breaks your text mode
1355[17:37:34] <Sentinel> I'm not convinced that's happening
1356[17:38:07] <jelly> Sentinel, those VTs might still be there logically. Go to Ctrl-Alt-F3, press enter a couple times, try to login with your username and password blindly
1357[17:38:25] <jelly> then go back to X and see if you're logged in
1358[17:38:26] <greycat> dka-: instead of assuming, read the release notes
1359[17:38:52] <dka-> Ok, I am on it: replaced-url
1360[17:38:52] <Sentinel> hah yeah I am
1361[17:38:55] <jelly> Sentinel, that's basically what would happen when I'd used nvidia
1362[17:38:56] <Sentinel> good trick jelly
1363[17:39:09] <jelly> it's just the video output is broken
1369[17:40:21] <greycat> I'll have to try to remember this one. "Physical problem such as aging hardware, loose cable, lighting in the room changed, flaw in your eyes; or you're trying to use nvidia on the Linux console"
1370[17:40:26] <jelly> noone can see what you're doing
1371[17:40:53] <Sentinel> yeah I just tell everyone I'm using a really good privacy filter
1372[17:41:00] <Sentinel> or my special glasses let me see
1373[17:41:18] *** Quits: Haudegen (~quassel@replaced-ip) (Quit: Bin weg.)
1392[17:52:16] <greycat> it's over 20 years old now, so I figure it's safe to spoil minor things like that especially when I don't name the novel or the author
1395[17:53:18] <shtrb> Is there some inherited problem to use at the same time an rfcomm connection for ip communication over bluetooth (DUN) and passing audio via HPF(?) to pulseaudio ? If I connect with rfcomm pulseaudio no longer play audio from my device.
1404[17:57:11] <shtrb> jelly, when even I connect over rfcomm there is no longer any audio output with pulse audio.
1405[17:57:20] <dka-> I am on step 4.2 of the release not for stretch, this is the output. How can I safely remove linux-image-3.16.0-4-amd64 and use the default kernel ? What is the kernel location ?
1417[18:00:06] <jelly> first make sure you a) have access to the console and a rescue image in case things go wrong b) look at which packages bring these files: dpkg -S /boot/vmlinuz*
1419[18:00:32] *** Quits: rgr (~user@replaced-ip) (Remote host closed the connection)
1420[18:00:44] <dka-> a) yes, all Dedicated have a rescue mode
1421[18:00:51] <jelly> perhaps you should patch-up that jessie first.
1422[18:01:02] <dka-> define "patch-up"
1423[18:01:13] *** Quits: fflori (~fflori@replaced-ip) (Remote host closed the connection)
1424[18:01:42] <jelly> apply all the missing upgrades for jessie up to this point in time. Then reboot it, see that things still work, then continue with jessie->stretch
1451[18:06:50] <jelly> dka-, if that's an option then great. Run a dist-upgrade as well. That ought to bring your jessie up to date.
1452[18:07:49] <jelly> since jessie is in ELTS, the only still supported kernel package is backported 4.9, so you could install linux-image-4.9-amd64 and reboot into that one.
1456[18:09:25] <jelly> also, you'll have a kernel to try that is slightly closer to the custom 4.11 you're running right now instead of old jessie 3.16.0-11-amd64 or whatever was the last ABI for jessie
1473[18:16:44] <Ether_Man> ;_; And the kernel-image package from experimental is not a drop in replacement for any other branch?
1474[18:17:45] <jelly> Ether_Man, if you're careful you can install a _kernel_ package on most branches, but in general packages in experimental are intended only for trying them out on sid (unstable)
1492[18:21:40] <jelly> Ether_Man, honestly that would be much easier. If you can find any d-i that boots and manages to install on your platform, you can install a newer kernel or any other package later.
1493[18:22:17] <Ether_Man> Installing is no problem. It's that network driver that kind of makes the system impossible to use since it's the only network it has :)
1515[18:28:58] <Ether_Man> No but it's a feature that carriers decides if it's available or not. Mine does not
1516[18:29:01] *** debhelper sets mode: +l 1184
1517[18:29:18] <hansol> why should i leave arch and move to debian ?
1518[18:29:34] <jelly> !why debian
1519[18:29:34] <dpkg> Debian strives to maintain your freedom whilst also paying close attention to the technical aspects of making a great OS. Debian is stable, upgradable and well tested. See also replaced-url
1520[18:29:42] <jelly> !why not debian
1521[18:29:49] <jelly> *cough*
1522[18:29:50] <jelly> !why not debian
1523[18:30:01] <jelly> dpkg, come on now!
1524[18:30:02] <dpkg> jelly: KCI error, or a problem with the Keyboard-Chair Interface.
1525[18:31:01] <jelly> hansol, Arch Linux is just fine. If it suits your purposes and serves you and your users well, no need to move away from it
1526[18:31:43] <Ether_Man> Are the kernel modules part of the kernel-image package? I assume they're their own package but named what?
1528[18:31:52] <greycat> dpkg, why not debian is <reply>Reasons not to use Debian may include the need for cutting-edge kernels/drivers for new hardware, or <SNS>.
1534[18:32:12] <dpkg> Factoid search of 'why not' by key (15 of 27): #debian why not ;; _default why not ;; why not ;; why not amd64 ;; why not aptitude ;; why not debian ;; why not dist upgrade ;; why not dist-upgrade ;; why not dmo ;; why not nvidia installer ;; why not nvidia-installer ;; why not pptp ;; why not sav ;; why not testing ;; why not use a ppa.
1535[18:32:32] <nkuttler> Ether_Man: they are part of the kernel image
1536[18:33:03] *** Quits: Sierra (~sierrakom@replaced-ip) (Remote host closed the connection)
1537[18:33:22] *** Quits: xsisec_ (~xsisec@replaced-ip) (Remote host closed the connection)
1551[18:40:20] <BlackSalmon> "gpg: absolutely no trusted keys" error (translation may not be precise), how can i fix that?
1552[18:40:46] <jelly> BlackSalmon, which debian release are you using?
1553[18:40:54] <jelly> !localized errors
1554[18:40:55] <dpkg> To provide command output in English instead of your native language, set your locale to an English one (e.g. C) prior to running the command, e.g. "LC_ALL=C apt-get -f install".
1572[18:47:02] * dpkg points at Ether_Man and laughs uproariously
1573[18:47:13] <Ether_Man> Would debian-10.6.0-amd64-DVD-1.iso have all the packages I need to compile my own or would I need to download more, beyond the kernel source that is ofc?
1574[18:47:33] <jelly> Ether_Man, if it's in source for 5.9, file a bug report against the package and ask the relevant CONFIG_ be enabled
1582[18:48:45] <judd> No packages in sid/amd64 were found with that file.
1583[18:48:45] <greycat> !dvd contents
1584[18:48:46] <dpkg> To find out which CD or DVD image contains a particular package or source file, use the search engine at replaced-url
1585[18:49:02] *** debhelper sets mode: +l 1178
1586[18:49:15] <jelly> well that's one point for ubuntu, then. I have this on 18.04: /lib/modules/5.4.0-48-generic/kernel/drivers/net/wireless/realtek/rtw88/rtw88.ko
1603[18:54:51] <BlackSalmon> I doesn't remember that i've installed that package
1604[18:55:22] <BlackSalmon> greycat: well, i thought that "sudo" just gives more permissions for command
1605[18:55:26] <greycat> the sudo man page actually claims that you can pass variables by using VAR=value but it doesn't say *where* to put that, and it doesn't show any examples, so who knows
1607[18:55:58] <sney> probably before the command, like with anything else. VAR=value sudo ...
1608[18:56:16] <greycat> that would be passing a variable to sudo, which it would probably then strip out of the environment
1609[18:56:32] <sney> this seems like a TIAS situation
1610[18:56:58] <greycat> sudo FOO=bar env | grep FOO= actually works for me, but I don't know what config options would cause specific variables to be kept or nuked
1611[18:57:23] <greycat> anyway, sudo env LC_ALL=C your command here is the most likely to work in any situation
1621[19:01:26] <BlackSalmon> at least `LANG` variable, `LC_ALL` is set correctly
1622[19:01:31] <greycat> ...
1623[19:01:53] <greycat> THE ENTIRE FREAKING PURPOSE OF INCLUDING LC_ALL=C INSIDE OF A COMMAND LIKE THIS IS THAT IT'S ONLY SET FOR THE DURATION OF THAT ONE COMMAND
1624[19:02:10] <greycat> Now, WHAT ERRORS DID YOU GET
1625[19:02:23] <BlackSalmon> oh, so i should include that IN command, not as a seperate one
1626[19:02:25] <BlackSalmon> D:
1627[19:03:07] <BlackSalmon> "no ultimately trusted keys found"
1628[19:03:47] *** Joins: magyar (~magyar@replaced-ip)
1639[19:14:34] <dpkg> [Basic Apt* Troubleshooting]. To diagnose your problem, we need ALL OF THE FOLLOWING information: 1. complete output of your apt-get/apt/aptitude run (including the command used) 2. output from "apt-cache policy pkg1 pkg2..." for ALL packages mentioned ANYWHERE in the problem, and 3. "apt-cache policy". Use replaced-url
1674[19:30:23] <greycat> It's your computer, so only you know what happened. Would you like us to guess?
1675[19:30:49] <genr8_> you added ubuntu to apt, it overwrote your debian trusted keyring with its own, now debian cant authenticate with its own servers
1678[19:31:16] <genr8_> you can try to fix it, but other packages probably got ruined
1679[19:31:21] <greycat> That's one guess. Another one is that this computer was installed as Ubuntu, and then you thought you could convert it to Debian.
1682[19:32:13] <BlackSalmon> It has to be debian from beginning, but it was my second approach to install linux so it could go wrong
1683[19:32:23] <mutante> remember there was once that script "debtakeover" or so? it was made for downloading it on some VPS with another OS and well.. "take over"
1684[19:32:32] <BlackSalmon> and it was few years ago
1685[19:33:01] <greycat> Your best bet would be a clean install. If you aren't willing to do that, you could try downloading the stretch version of the package and installing that by hand.
1686[19:33:38] <BlackSalmon> Nah, if I have to spend another 30 minutes and not fix this imma done with this version
1687[19:34:13] <BlackSalmon> I was just sticking it because i didn't wanted to move my code to newer version of debian. Looks like now i have to do that
1732[20:04:35] <sponix2ipfw> dka: curl is version 4 in Buster
1733[20:06:59] <alex11> jessie is also obsolete so i'm sure you've heard this before but consider updating to at least stretch and then preferably buster... unless this is an offline machine
1768[20:25:34] <NetTerminalGene> in debian developer tracker stuff it says "Migration status for thunderbird (1:68.12.0-1 to 1:78.3.1-2): BLOCKED: Rejected/violates migration policy/introduces a regression" does it mean it will never be released?
1769[20:26:28] <genr8_> no, it means devs get to fight about why and what to do about it
1770[20:26:42] <NetTerminalGene> genr8_: i was just changing its names on archlinux
1791[20:41:45] <dvs> dka-: ??? But the latest is bullseye??
1792[20:41:52] <greycat> No.
1793[20:41:56] <greycat> !buster
1794[20:41:56] <dpkg> Buster is the codename for the current <stable> release, Debian 10, released 2019-07-06. "Buster" is Andy's pet Dachshund in Toy Story, see replaced-url
1820[20:55:37] <genr8_> so most of the options are set from within systemd then
1821[20:57:25] <dka-> I did apt-get install linux-image-4.9-amd64, how do I restart and use that kernel instead of the current one?
1822[20:57:32] <Ether_Man> aaand ofc... not all packages to compile kernel was on dvd1... Unable to locate package libelf-dev... Because that's on dvd2... sigh...
1828[20:58:59] <genr8_> grub is what controls booting into kernels. if you installed that package, it should have re-ran and configured grub so theres an option listed. reboot, then Go to the advanced menu
1829[20:59:11] <genr8_> if its a lesser version, its probably gonna not be the default.
1830[20:59:33] <dka-> It's a dedicated server, I can't choose in grub the kernel to use
1831[20:59:44] <dka-> I have to wait until SSH service is up, after grub
1850[21:03:24] <greycat> if there's a way to do it from /etc/default/grub I have never seen it...
1851[21:03:24] <genr8_> I have a guess that its the GRUB_DEFAULT number
1852[21:03:33] <genr8_> 0 is "first boot option" , 1 is second, and so on
1853[21:03:45] <dvs> !price genr8_
1854[21:03:50] <dvs> heh
1855[21:04:02] <genr8_> except theres also "Rescue" modes. so 1 is the rescue mode of #1. i believe 2 would be the first second kernel.
1856[21:04:13] <genr8_> er 1 is the rescue of 0
1857[21:05:08] <sigint> In any case, you can inspect /boot/grub/grub.cfg once you've regenerated your grub conf to know for sure
1858[21:05:45] *** Quits: derpadmin (~derpadmin@replaced-ip) (Remote host closed the connection)
1859[21:06:53] <dka-> What do you mean "regenerate" ? How do I "regenerate" my grub conf? this is my /boot/grub/grub/cfg => replaced-url
1860[21:07:05] <dvs> dka-: update-grub
1861[21:07:08] <greycat> the other thing you can do is put a script in /etc/grub.d/ which has a very low number, so that it puts its payload into the generated menu before the 10_linux script does
1862[21:07:39] *** Quits: chele (~chele@replaced-ip) (Remote host closed the connection)
1863[21:07:40] <genr8_> scroll down halfway while reading that file, the first option is 'Loading Linux 4.9.0-0.bpo.12-amd64'
1864[21:08:07] <genr8_> in fact 4.11 is not listed there at all
1865[21:08:22] <genr8_> oh nevermind i missed it
1866[21:08:36] <genr8_> its shown above it in its own /etc/grub.d/06_OVHkernel section
1867[21:09:01] *** debhelper sets mode: +l 1176
1868[21:09:07] <greycat> I see no documentation for the GRUB_DEFAULT variable in /etc/default/grub, and no obvious way to match each "menuentry" (if that is even the thing one should be counting!) to a numeric value
1869[21:09:07] <genr8_> In that case, try GRUB_DEFAULT=1
1876[21:11:03] <dka-> This is my configuration: replaced-url
1877[21:11:47] *** datasmurf2 is now known as datasmurf
1878[21:12:10] <genr8_> it worked
1879[21:12:24] <greycat> Oh lovely, the documentation talks about "menuentry ... --id ..." but there are no such --id flags in the Debian grub.cfg file.
1882[21:13:17] <dka-> I have another host that didn't work to update-grub, it has an extra kernel 3.16 so I did apt-get remove linux-image-3.16.0-4-amd64 but update-grub still show the old kernel, this is the output: replaced-url
1883[21:13:32] <dka-> how do I correctly remove the unused and old kernel 3.16?
1887[21:14:28] <dvs> dka-: you did get rid of it. however, there are two of them.
1888[21:14:50] <genr8_> yeah theres 0-4 and 0-11
1889[21:15:00] <dka-> I did not get ride of : Found linux image: /boot/vmlinuz-3.16.0-11-amd64 and Found initrd image: /boot/initrd.img-3.16.0-11-amd64
1890[21:15:13] <genr8_> you can apt remove that one next
1891[21:15:26] <dka-> I already did, with apt-get remove
1909[21:18:02] <genr8_> you would wanna reboot into the new kernel first. so it knows not to get overzealous and try to remove that
1910[21:18:04] <dka-> It's not, it's Linux master-bhs-01 4.11.0-xxxx-std-ipv6-64 #1 SMP Mon May 8 14:47:33 CEST 2017 x86_64 GNU/Linux
1911[21:18:13] <genr8_> hmm
1912[21:18:18] <greycat> but that's a non-Debian kernel.
1913[21:18:29] <dka-> Yeah, I have 3.16.0-11 to remove, and I need to downgrade from 4.11 to 4.9
1914[21:18:32] <greycat> your last *Debian* kernel is the linux-image-3.16.0-11-amd64 one
1915[21:18:39] <genr8_> ah
1916[21:19:00] <dka-> greycat, why do you say that? to me it's 4.11, I don't remember using 3.16.0 at any time
1917[21:19:14] <sigint> dka-, looking at the output of update-grub, you really just need to do that: replaced-url
1918[21:19:16] <dka-> it's a mistake if it's stil there
1919[21:19:19] <greycat> you're still on jessie, so linux-image-amd64 points to linux-image-3.16.0-11-amd64
1920[21:19:37] <dka-> greycat, if I remove linux-image-amd64, then it will break my OS
1921[21:19:42] <greycat> *sigh*
1922[21:19:48] <dka-> it won't?
1923[21:19:49] <greycat> then don't fucking do it
1924[21:19:55] <dka-> ok, I just want to keep cleans
1925[21:20:02] <greycat> I don't even care any more. Do whatever the hell you want.
1926[21:20:03] <dka-> so I'll do what sigint tells me and reboot
1927[21:20:27] *** Quits: soul-d (~name@replaced-ip) (Remote host closed the connection)
1928[21:20:36] <dka-> greycat, excuse me, was I rought with you? I didn't meant to
1929[21:20:36] <genr8_> dont overreact
1930[21:21:16] <genr8_> you can actually remove that package, its more of a warning you're going about the wrong way
1931[21:21:17] <dka-> It's a stressful step to upgrade jessie to stretch, I am trying to get the proper information, myself is not enough
1932[21:21:50] <greycat> linux-image-amd64 is just a metapackage that points to the current stable kernel for your release, it doesn't actually DO anything except point to a real kernel version
1933[21:22:23] <sigint> dka-, let's focus. You currently run an OVH kernel installed outside of apt. It is still the default used by grub because of the /etc/grub/06_OVHkernel
1934[21:22:44] <sigint> You have multiple Debian kernels installed, but they won't be used. You can keep them, don't remove them
1935[21:23:17] <sigint> by changing the order of 06_OVHkernel, you will put your current OVH kernel behind your latest Debian kernel
1936[21:23:22] <dka> I have 3 host that needs to upgrade, all part of a mesos network, it's 3 master, I am updating them 1 by 1, on 1 onf them , I did mv 06_OVHkernel 96_OVHkernel and I am rebooting to see if I can get to 4.9
1937[21:23:41] <sigint> Did you run update-grub?
1938[21:23:42] <genr8_> you have to rename, update grub, reboot.
1939[21:24:11] <genr8_> I still like my idea of changing GRUB_DEFAULT=1 in /etc/default/grub
1940[21:24:13] <dka> yes, this is the output of update-grub : replaced-url
1941[21:24:30] <dka> as you can see, 4.9 is listed first , so it should pick it
1942[21:24:32] <dvs> grub.cfg would be better
1943[21:24:36] <greycat> after update-grub you can read the generated file /boot/grub/grub.cfg and see where it put each kernel in the menu
1944[21:25:05] <sigint> dka, perfect IMO you're good to go
1945[21:25:07] <dka> both options seems to do the exact same thing. I don't see any human using the grub order anyway
1946[21:25:22] <genr8_> reboot and call it a day.
1947[21:25:33] <dka> yeah, the host does not start the openssh-server ^^
1952[21:26:39] <genr8_> was that really required first ?
1953[21:26:39] <dka> I still have docker-ce installed, listed when I run apt-forktracer | sort , can I keep it during stretch upgrade or is it really advised to remove it first then reinstall ?
1954[21:27:03] <dka> genr8_, since I used to have a custom kernel, and I want the upgrade to be smooth, yes
1955[21:28:34] <genr8_> ,v docker-ce
1956[21:28:35] <judd> No package named 'docker-ce' was found in amd64.
1957[21:28:55] <sigint> dka, do you have any container running? Do you know how to recreate them if necessary?
1958[21:28:56] <genr8_> well theres no replacement, so either the old one works or it doesnt, and you fix that later
1959[21:29:12] <dka> sigint, yes I have about 50 containers running, and yes i know how to recreate them if necessary
1960[21:29:19] <sigint> genr8_, docker is probably installed outside of the Debian repos
1961[21:29:51] <dka> But my strategy is to upgrade jessie to stretch , then sretch to buster, have the 3 master keeping the state, then do it again on another host, and finally and the last host
1962[21:30:04] <dka> this should cause no downtime since there are always 2 master up, and the state can be kept
1963[21:30:19] <dka> sigint, yes it is installed from the docker repo
1965[21:30:48] <dka> that's why I am thinking to keep it installed while doing the upgrade to stretch, so I can reconnect the 3 masters together and see that the upgrade went well on the first step
1966[21:31:04] <sigint> Assuming all the containers use volumes you should be able to stop them, uninstall docker, upgrade Debian, reinstall docker (the docker repo URL may have changed), restart containers
1967[21:31:27] <genr8_> well yeah, i just meant its not gonna be able to auto update, so either the old one just works with the new OS, or it doesnt and you have to fix it
1970[21:32:48] <trysten> is there a tool to debug permission errors? i'm trying to figure out why apache can't access a certain dir
1971[21:33:19] <genr8_> the human brain
1972[21:33:59] <trysten> yeah mine is not great. i've been using it to debug permission errors for years and it still sucks lol
1973[21:34:03] <sigint> dka, do you run the latest and greatest version of Docker? If so maybe keeping docker and updating in place will work. What does docker info say?
1974[21:34:17] <genr8_> you want the directories to have the "rx" permission set for the user/group of apache, which is likely replaced-url
1975[21:34:48] <dka> sigint, it says this initially : replaced-url
1976[21:34:54] <genr8_> or "rx" (5) set for "other" (3rd slot)
1977[21:35:04] <dka> The host that has reboot on kernel 4.9 is still hanging,(docker not started)
1992[21:42:16] <genr8_> trysten, you're on debian right ? on other OS'es Apache is locked down by a security mechanism called SELinux.
1993[21:42:39] <sigint> dka, I'll be honest with you: I don't understand why you decided to upgrade these machines. As you run containers and know how to recreate them you can use a fresh install of buster, install docker, copy your container data and be rolling
1994[21:42:39] <trysten> yeah no SELinux here.
1995[21:43:19] <genr8_> then its just regular permissions then. are you trying to access files in user's home directories ? cause thats confusing
1997[21:44:30] <genr8_> you should have a system dir for apache that has the right permissions like /var/replaced-url
1998[21:44:47] <dka> sigint, because the Host configuration script is old and got hacked by the time we installed the machine, redoing a fresh install script for the installing the Buster OS would be the right action, but I don't want to affoard the time now it will be quite expensive
1999[21:45:02] <dka> weird thing, after restarting the OS on 4.9, docker process never listen
2026[21:55:39] <Franciman> the disconnection occured around 21:00
2027[21:55:54] <Franciman> so it's at the middle of the logs
2028[21:56:19] <Franciman> there are 4 failures like this:
2029[21:56:21] <Franciman> ott 05 21:07:16 picci systemd-timesyncd[432]: Timed out waiting for reply from 5.249.149.117:123 (2.debian.pool.ntp.org).
2030[21:56:37] <Franciman> I don't know if this is the cause or the symptom of the disease affecting my connectivity
2031[21:56:58] <Franciman> then a huge amount of DHCPREQUESTs starts
2032[21:58:24] <dka> I am now on linux kernel 4.9 from debian repos, I think I am ready to jessie -> stretch
2033[21:58:25] *** Quits: BlackSalmon (05ada0ce@replaced-ip) (Remote host closed the connection)
2034[21:58:59] <dka> I think that's step 4.3 from release note
2035[21:59:46] *** Quits: dvs (~hibbard@replaced-ip) (Remote host closed the connection)
2036[21:59:48] <dka> Do I just run `sed -i "s/jessie/stretch/g" /etc/apt/source.list; apt-get update; apt-get upgrade; apt-get dist-upgrade` to upgrade to stretch ?
2037[21:59:56] <genr8_> yes
2038[22:00:10] <greycat> you misspelled sources.list
2039[22:00:33] <jmcnaught> Maybe consider running inside a screen session, there are other suggested steps in the release notes instructions
2040[22:00:41] <dka> and then, I think I will restart to see if the mesos master reconnect well, and then, I can do the same to upgrade to buster ?
2041[22:01:42] <greycat> dpkg, stretch->buster
2042[22:01:42] <dpkg> Read (at least) the upgrading chapter of the <release notes> replaced-url
2043[22:02:08] *** Quits: ledeni (~ledeni@replaced-ip) (Remote host closed the connection)
2045[22:03:19] <jmcnaught> I don't know why anyone would want to upgrade without knowing what's new and checking the issues to be aware of.
2046[22:03:24] <dka> I wass missing the apt full-upgrade thx
2047[22:03:40] <dka> jmcnaught, for the record, jelly already gaved me that link
2048[22:03:46] <dka> I am already following it
2049[22:04:18] <genr8_> Franciman, the answer is WiFi sucks
2050[22:04:43] <dka> also, I used buster and stretch on my laptop, I know what changed in the tooling (like /etc/rc.local removed) and how to restore the services, I used docker for all of them, it's normally safe to upgrde
2051[22:05:13] <genr8_> apt full-upgrade is the same as apt-get dist-upgrade
2052[22:05:25] <Franciman> genr8_, lol, at least is there a way to restart the whole wifi?
2053[22:06:22] <genr8_> yea. perhaps through the rfkill interface
2054[22:06:40] *** Quits: OERIAS (~OERIAS@replaced-ip) (Remote host closed the connection)
2107[22:24:06] <Franciman> dotcom, never tried the wired ethernet, because the pc has only usb-c ports
2108[22:24:27] <dotcom> Franciman: subsequent questions, are there other wireless devices connected to the router? do they lose connectivity when your laptop does?
2109[22:24:29] <genr8_> have you tried the wifi adapter in another PC
2110[22:24:37] <Franciman> ok, no
2111[22:24:41] <Franciman> the other devices work fine
2112[22:25:13] <genr8_> what device is the problematic one anyway
2113[22:25:35] <Franciman> my dell xps 13 9380 laptop
2220[22:42:11] <dka> Is update-initramfs: Generating /boot/initrd.img-4.9.0-0.bpo.12-amd64 the last line of the upgrade
2221[22:42:12] <dka> ?
2222[22:42:59] <sney> dka: this looks fine, now it's time for 'apt-get dist-upgrade'
2223[22:43:49] <sney> if you were using unattended-upgrades, you may need to configure it again once the process is finished. but there are no errors here.
2224[22:43:55] *** Quits: Tom01 (~tom@replaced-ip) (Remote host closed the connection)
2238[22:50:42] <sney> wicd has a python2 dependency, and upstream is largely inactive, so you *might* need to find an alternative when you upgrade to bullseye.
2239[22:50:45] <sney> but you've got time.
2240[22:51:06] <ryouma> it didn't seem like it even applied to me
2273[23:05:35] <genr8_> most of this stuff seems like they are old versions that got deprecated and replaced by entirely new versions with the same but slightly different package names
2278[23:08:00] <judd> No package named 'libjasper1' was found in buster/amd64.
2279[23:08:12] <genr8_> <shrug>
2280[23:08:29] <genr8_> JasPer JPEG-2000 runtime library. JasPer is a collection of software (i.e., a library and application programs) for the coding and manipulation of images.
2292[23:18:09] <hexo_> hi everyone, is it possible to get virt-clone without installing stuff like libgtk-3 or anything in common with graphical ui? this is server with no screen at all so it makes no sense to have deps like that
2352[23:49:16] <dka> Ok, so upgrading jessie to stretch while running a mesos platform on 3 master nodes was a success and without any hassle, I am now looking to upgrade to buster, using the release note
2354[23:49:45] <sney> it will be basically the same process
2355[23:50:24] <greycat> probably easier because you've already cleaned up all the messy bits
2356[23:50:34] <jelly> you never said you were running a cluster
2357[23:50:57] <dka> the longest downtime was after switching to the default kernel instead of the ovh one, for a reason I don't know, I had to reboot an extra time, otherwise the downtime was about the time to reboot. Use mesos :)
2358[23:51:17] <jelly> or perhaps you did, I had some paid work to do instead of irc volunwastingtime
2359[23:51:48] <dka> no worries, I wish I could be in the same situation :)