180[02:11:17] <n4dir> there is nothing you could do about browsers. The ones which are lightweight are probably not what you want. The ones who do what you want are not lightweight. That would be either firefox or chromium. (the other ones dillo, netsurf, not in the repos anymore, and some text-based ones)
181[02:11:56] <amlchief> chromium has problems currently, a lot of crashes
182[02:12:10] <wallbroken> do you know iceweasel ?
183[02:12:11] <alex11> wallbroken, arrow up/down or middle click+mouse left/right
184[02:13:06] <alex11> this is also all explained in the man page
188[02:15:51] <wallbroken> i have a trackpad, not a mouse
189[02:15:54] <alex11> re browsers: i'd probably just use firefox, firefox-esr in debian if you're not picky or upstream if you want newer features+certain extension compatibility
190[02:16:03] <alex11> well, arrow keys will still work
191[02:16:08] <wallbroken> ok
192[02:16:22] <wallbroken> icewasel, do you know that?
196[02:18:31] <alex11> not in recent debian versions
197[02:18:34] <alex11> !iceweasel
198[02:18:34] <dpkg> Iceweasel was a DFSG-free fork of <firefox>, replacing it in Debian since 4.0 "Etch". See <why iceweasel>, <iceweasel vs firefox>, <iceweasel user-agent>, replaced-url
214[02:31:14] <mutante> trying to debug why a PXE install does not work. I have a DHCP server and see the DHPACK, i have a TFTP server and I see it "Serving lpxelinux.0", i have a HTTP server and i see the new host GETting /tftpboot/buster-installer/debian-installer/amd64.initrd.gz from it.. and that's where it ends. On my new machine I never see the Debian installer in the console. (It works for existing hosts but in other
215[02:31:20] <mutante> subnets). What is the next step after that that could fail?
216[02:33:07] <mutante> well, i see another DHCP ack on a different machine and then that's it. but that just adds to my confusion. I guess i am looking for a document that describes each step in the boot process until i should actually see the installer.
217[02:33:52] <mutante> or it's about some ACLs on network gear I would have to ask to be changed.. but since i already see all the protocols above.. not sure what I have to ask for
220[02:41:46] <kingsley> How would you extract the contents of a file 1.) whose name ends with a ".exe" suffix, and 2.) that is described by the venerable "file" command as "PE32 executable (GUI) Intel 80386, for MS Windows"?
222[02:43:53] *** Quits: electro33 (uid613@replaced-ip) (Quit: Connection closed for inactivity)
223[02:44:20] <mutante> kingsley: depends what you want to get out of it. do you want to extract images? you won't be able to just extract source code. it's a binary
224[02:45:08] <kingsley> mutante: Yes, your guess is correct. I expect the .exe file contains images.
225[02:45:43] *** Quits: pvdp66556 (~pvdp@replaced-ip) (Remote host closed the connection)
226[02:45:55] <mutante> kingsley: you can look at it with a hex editor, like hexer. I also remember this Windows program to extract images. replaced-url
227[02:46:35] <mutante> maybe reshacker.exe would run in wine
234[02:48:23] <kingsley> jmcnaught: No, I have not, yet.
235[02:48:43] <jmcnaught> "Note that self-extracting ZIP files are supported, as with any other ZIP archive; just specify the .exe suffix (if any) explicitly" -- unzip(1)
236[02:48:44] *** Quits: leorat (~leorat@replaced-ip) (Remote host closed the connection)
265[03:18:30] <Whyvn> I tried to upgrade my packages, and there are 2 packages I cannot upgrade. When I do apt update and apt upgrade, I see 0 upgraded, 0 newly installed, 0 to remove and 2 not upgraded.
357[04:34:21] <willma> I was installing build-essential on old system but there has unmet dependencies...When I use aptitude, I found that it could solve the problem by downgrade one package and it worked very well for me ... Now the problem is that I want aptitude to automatically respond to choose the solution of the resolver since I was building a Docker image ... Does anyone know how to use aptitude to automatically choose the solution ???
521[08:09:30] <wowas> i installed debian with the installer on md0 (software raid). How ever, if i unplug the first disk, the machine wont boot. It does not even load grub.
522[08:09:35] <wowas> any hints about this?
523[08:10:37] *** Quits: citypw (~citypw@replaced-ip) (Remote host closed the connection)
529[08:16:50] <jelly> wowas, you probably need a boot loader on the second disk. If the machine is booting in UEFI mode you also probably need a second EFI system partition (and it's not a good idea to have two of those, I hear)
539[08:21:49] <jelly> it mirrors the parts that are mirror members. If you only have two disks and they're bootable, that's probably just two matching partitions and not the whole disks
543[08:23:41] <jelly> so in general, not whole disks, no. You can do it that way but then it's very hard to also keep one or both of those disks bootable
559[08:32:03] <dominix> I have tried a tcpdump with the address on the other ipv4 ... and nothing
560[08:32:35] <foul_owl_> What is a good alarm clock program for debian? There used to be one I used for jessie but after upgrading sadly it seems to be gone
561[08:32:47] <dominix> my provider is huricane electric
633[09:32:27] <kassle> hi, i'm building unattended debian cd. using simple-cdd. anyone know how to strip down debian-installer and keep the text based installer only ?
699[10:49:00] <oxek> why does the default PATH on debian include both /usr/bin and /bin? They point to the same place. Doesn't that mean that command lookups happen twice?
747[11:17:57] <blebz> Does anyone know where the config files are for Apache on a default install of buster... I want to change the folder that Apache is looking for for .html files etc.
748[11:18:10] <blebz> Whats the best way to re-config Apache..
749[11:18:28] <blebz> I asked it to be installed when I set-up server...
750[11:18:33] *** Quits: kreyren (~kreyren@replaced-ip) (Remote host closed the connection)
768[11:25:34] <jelly> and in general, if there's a README.Debian it's a good idea to read it
769[11:25:37] <jelly> !readme.debian
770[11:25:37] <dpkg> README.Debian (or README.Debian.gz) is a document found in the /usr/share/doc/$packagename/ which explains any Debian specific details in the package's operation or configuration.
784[11:36:16] <blebz> Hmm.. the thing is though I want to change the document root cos I physically want to have my Webdocz on a different HD (raid) not on the system disk.
785[11:37:07] <blebz> I'm starting to think maybe I should have said no when installing Debian (to installing webserver) then maybe I would have had more control over it's config..
901[13:38:04] <bs0d> so these snaps and flatpacks - are they alternatives to apt? I've never used those, but I will google if I can use them to get Thunderbird
902[13:38:55] <ratrace> they're containerized application package managers. a way to install software as pre-made containers so it doesn't mess with the rest of your system
903[13:39:21] <ratrace> there's a snap of thunderbird, but it's not official. caveat emptor. dunno about flatpak
904[13:40:11] <ratrace> snaps and flaptaks are their own "stores" or hubs, kinda like dockerhub. it's software packaged outside of debian repos, outside of debian QA and testing.
936[14:05:03] <ratrace> tartar: by finding the cause of the freeze. is it just temporariy unresponsiveness, or hard lock. is there anythign in the dmesg or journal when that happens.
940[14:05:54] <tartar> ratrace, im not entirely sure. dmesg shows some warnings like CPU11: Package temperature above threshold, cpu clock throttled (total events = 1)
941[14:05:55] <ratrace> try tailing dmesg (with -T -w, not `tail`) while you build and see if it'll spit anything out. also, is that perchance the first gen and zen1 cpu?
942[14:05:58] <ratrace> *amd
943[14:06:11] <tartar> it's an i7-8700k
944[14:06:52] <ratrace> core i7 has a treshold at 100°C, can't remember if it's hard lock or temporary lock, or just throttling
945[14:07:17] <ratrace> but sure, check the temps. maybe the cooler needs undusting, and/or thermal paste reapplied. I do that yearly on my computers.
980[14:27:16] <ratrace> are you sure there's no dust? it's near impossible to have no dust at all. did you actually remove the cooler to clean it without reapplying the paste?
981[14:27:23] *** Quits: horner (~horner@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
982[14:27:24] <jelly> you probably want to avoid running at 90C for months, my i3 CPU got unstable after that and only worked stable if just a single core was running
983[14:27:45] <tartar> i didn't remove the cooler
984[14:27:56] <ratrace> then you probably didn't undust it properly.
985[14:28:12] <tartar> jelly, that's just for some minutes when compiling something
986[14:28:21] <tartar> ratrace, i'
987[14:28:31] <tartar> i've seen dusty pcs, this isn't dusty
989[14:28:58] <jelly> tartar, oh, what's the idle temp like? More than 40-45C?
990[14:29:26] <ratrace> even if you don't see (much) dust inside, it'll get stuck in the very tiny spaces between cooler fins and you can't see those until you remove the fan, remove the cooler and look with a light background
991[14:29:33] <tartar> jelly, about 40, and probably lower (it's been at around 100 for some minutes now, so it'll take a few minutes to cool)
992[14:29:49] <jelly> that sounds okay
993[14:30:00] <tartar> ratrace, and THAT would make it overheat to 100c and throttle?
994[14:30:09] <ratrace> except the 100°C part. under load, i7 shouldn't get over ~90 under normal conditions
997[14:35:40] <certaindestiny> Hi all, i am having a small issue with udhcpc on Kali, I asked on the kali IRC and they told me to come here since the package orginates here. currently whenever i try to get a default route on WWAN0 using udhcpc -i wwan0 i do not get the default route pushed if ETH0 is connected.
998[14:35:59] <certaindestiny> i can see in the DHCP packets that the route (option 3) is being sent by the DHCP server
1001[14:38:44] <certaindestiny> ksk, i asked on kali-linux and they told me to ask here if it is a known issue. If you do not know or feel like it is not the correct place to ask please do tel me in a more polite matter than calling kali.
1002[14:39:10] <ratrace> there's nothing impolite here
1003[14:39:12] <ratrace> !based on debian
1004[14:39:12] <dpkg> Your distribution may be based on and have software in common with Debian, but it is not Debian. We don't and cannot know what changes were made by your distribution (compare replaced-url
1005[14:39:16] <jelly> certaindestiny, our channel and #kali-linux channel have incompatible policies.
1008[14:40:50] <jelly> certaindestiny, they say anything that is imported unchanged from Debian should be solved in Debian; we say we don't deal with _ANY_ issue in a derivative distro, because it's impossible for us to know the interactions between changed and unchanged parts and differences in the environment.
1009[14:41:23] <certaindestiny> ratrace. The polite way would have been to to provide some text and explanation instead of only calling kali, i do not disagree with the policy i disagree with the way it is communicated to me
1010[14:41:40] <jelly> both points of view have some merit, but the end result is you get help nowhere
1011[14:42:02] <jelly> certaindestiny, if you can reproduce the same issue with an actual Debian installation we can try to help.
1012[14:42:04] <ratrace> certaindestiny: but you're not the first nor the last to ask about derivatives. we just grew too tired typing up same things over and over again, hence the factoid
1013[14:43:18] <jelly> if it were isc-dhcp-client I'd say look at the config file
1014[14:43:42] <ratrace> besides, the factoid nicely and politely explains we can't provide support and gives alternative means for kali. if that's not polite, what is? a perfumed letter in gold framed stationery delivered by a penguin cosplayer?
1015[14:43:51] <ksk> certaindestiny: while I partly do unsterstand what you are trying to say, please also understand me/us: People do ask such questions regularly, and I dont really want to write my own text containing the same information anytime.
1018[14:44:40] <certaindestiny> jelly, thanks. seems like i am once again in no-mans land.i do not think udhcpc is based in the isc client version but perhaps somebody knows
1021[14:45:42] <jelly> certaindestiny, ##linux channel will however accept questions for any distro
1022[14:45:43] <certaindestiny> ratrace, The polite thing would have been to let me ask in the off change that somebody in this IRC has ran into the same issue either debian or another distro. Only stating the factoid does not want me to be a part of this IRC.
1023[14:46:41] <ratrace> so I guess you don't understand what the factoid is saying.
1024[14:47:03] <ratrace> the part about how we can't know what derivatives do, and thus can't offer a solution
1035[14:49:13] <certaindestiny> ratrace, There is no way everbody in this IRC only has experience with debian and no other software at all. As it is related there is chance of somebody knowing the answer and since i am finding now help except a bug from 2013 i am going to ask.
1036[14:49:27] *** Quits: cupcake90 (~cupcake90@replaced-ip) (Remote host closed the connection)
1037[14:49:31] <certaindestiny> no*
1038[14:49:43] *** Quits: Sciuro (~screwy@replaced-ip) (Quit: No nuts, no glory!)
1043[14:50:09] <ksk> certaindestiny: and as someone stated, for example the ##linux channel on this network would be suitable for such an approach. It is not in here.
1044[14:50:10] <Haohmaru> and there are a ton of people on it
1048[14:50:46] <ratrace> or else let's just invite the entire #ubuntu here, I'm sure they'd appreciate the expertise.
1049[14:50:52] <jelly> certaindestiny, we actively enforce the topic. If said "somebody" knows the answer they can talk to you in some _other_ channel or in PM, not here
1050[14:50:55] <Haohmaru> certaindestiny even "debian testing" (that is the next version) is NOT supported here, only the stable (and previous stable) version
1051[14:51:28] <jelly> people on irc can join multiple channels at the same time
1053[14:51:55] <Haohmaru> and multiple networks.. debian-next has a channel on another network
1054[14:51:58] <ksk> certaindestiny: a workaround could certainly be: Install Debian Buster, reproduce the Issue -> report back to us (or the bugtracker, or ..)
1055[14:52:08] <certaindestiny> the more people the more chance. and i am not going to ask 50 million times. i asked once and i understand the policy, however i do not like the response i got. thats all
1061[14:53:42] <Haohmaru> the first debian i installed was debian-next, when i quickly got stuck, i came here to ask for help, i was also immediately told to GTFO and go to the debian-next channel on OFTC, i didn't like it
1062[14:54:08] <Haohmaru> when it broke, i installed debian stable, and now i am allowed to be here ;P~
1063[14:54:28] <ratrace> I still don't see what's wrong with the factoid. it points at #kali-linux, at their forum, points at more factoids explaining why this is a problem, and a guide on asking questions. is there something missing?
1064[14:54:39] <certaindestiny> i am keeping with the. if dont have anything polite to say, say nothing at all.
1065[14:55:03] <greycat> We can definitely arrange that you can say nothing at all.
1066[14:55:22] <ratrace> lmao
1067[14:55:40] <Haohmaru> ratrace i'd think people would first go in the linux channel, and/or maybe in their distribution's channel, ask there, listen to the <crickets.wav> for hours, get semi-pissed off and start asking in any "linux-like" channel until they end up here, and here it's very alive
1074[14:56:31] <certaindestiny> ratrace, the problem is not with the factoid but the way it is presented. if i am not allowed to ask if this is a know issue on a upstream distro how am i supposed to find out
1077[14:56:45] <dpkg> Bug Tracking System for Debian packages, replaced-url
1078[14:56:56] <greycat> If it's a known issue in a Debian package, there should be a bug for it.
1079[14:58:11] <ratrace> certaindestiny: or you could just say "Yaeh, I know this is Kali, but anyone knows if this is a problem on Debian?" and save a lot of kilojoules whining whether raising a factoid is impolite... nobody said anything else, or was abrasive, and the factoid itself is polite and informative.
1080[14:58:21] <jelly> ratrace, you're missing the fact that #kali-linux will actively punt a user back here if they think the problem is with a piece they imported from Debian as-is
1081[14:58:21] <certaindestiny> should and are, different worlds.
1096[15:01:47] <ratrace> certaindestiny: sorry, I don't see you specifically stating, along the lines of, "I know this is kali, but is this an issue on debian stable too?". your question implied you're asking because "kali is just debian" as a lot of people tend to think, hence the factoid. I mean it's not your fault, it's just that a LOT of (skiddie) folks comes to ask about Kali issues. hence the trigger
1099[15:02:14] <greycat> "I'm using Kail version x. I have a problem with pkg y. I could not find a Debian bug for my issue, so I installed Debian 10 and reproduced the issue. Here are the symptoms: ____"
1104[15:03:01] <Haohmaru> "but, but, but.. Ubuntu is debian"
1105[15:03:02] <greycat> Kali, Ubuntu, MX and Parrot are the biggest offenders.
1106[15:03:36] <ratrace> even in that order of frequency, lol.
1107[15:03:37] <TuxCrazy> greycat, why? why do you call them offenders?
1108[15:03:43] <certaindestiny> ratrace, Like i said. since debian is upstream that is logical thing to do. But getting a come back where you came from wench response withing a couple of seconds is not they way i would welcome people to my IRC. even if they are way out of topic
1109[15:03:46] <greycat> Because they offend us.
1110[15:04:01] <ratrace> we're.... triggered :)
1111[15:04:10] <Haohmaru> TuxCrazy they waste lots of times often, by hiding (even actively) the fact that they aren't actuall running Debian stable
1112[15:04:38] <certaindestiny> haohmaru, Next time ill try that ;)
1113[15:04:53] *** ChanServ sets mode: +o greycat
1114[15:04:54] <TuxCrazy> Haohmaru, distros like MX Linux aren't running Debian stable?
1115[15:04:55] <Haohmaru> and then, you will have zero right to say that you've been offended
1122[15:05:44] <greycat> TuxCrazy: Correct. MX Linux is NOT debian and is NOT supported here. MX Linux questions are NOT on topic here, and we will NOT answer them.
1123[15:06:13] <Haohmaru> TuxCrazy if it's not called "Debian stable" and doesn't come from debian.org or its mirrors, it's probably not supported here
1124[15:06:22] <Haohmaru> that includes "Debian Next" even
1125[15:06:35] <TuxCrazy> ok
1126[15:06:48] <Haohmaru> even tho, after some time, Next will become Stable
1127[15:06:56] <TuxCrazy> can we use Debian backports on Debian stable?
1128[15:07:00] <Haohmaru> yes
1129[15:07:11] <greycat> When testing is fully frozen, we tend to support it here, briefly.
1130[15:07:12] <TuxCrazy> what are the risks associated with that?
1131[15:07:16] <ratrace> it's normal for various software/community/distro components to have dedicated channels. cuts down the noise
1132[15:07:45] <greycat> TuxCrazy: the backported packages are basically snapshots of a non-stable package at a particular moment in time. It could have bugs, and those bugs might not be fixed any time soon.
1133[15:07:55] <Haohmaru> greycat i was roughly trying to show that it's not like we are specifically pissing off other distros
1134[15:08:04] <Haohmaru> and that the topicness is serious
1143[15:11:08] <greycat> It's also worth noting that the presence of a backported package can sometimes cause issues during a release upgrade. Be prepared to remove (temporarily or permanently) any backported packages when going from Debian 10 to 11, etc.
1144[15:11:50] <Haohmaru> hmz
1145[15:12:01] <greycat> certaindestiny: you were muted because you threatened to come back in the future and pretend to be a Debian user to get help with your non-Debian operating system
1146[15:12:05] <Haohmaru> is debian 11 coming soon?
1147[15:12:15] <jelly> !wwbr
1148[15:12:15] <dpkg> from memory, wwbr is now Debian "Bullseye" is the current testing branch as of 2019-07-06 and it will be released "when it's ready."
1149[15:12:16] <greycat> Not until next year, at the earliest.
1150[15:12:34] <greycat> There's a partial timeline on the wiki.
1151[15:12:38] <Haohmaru> okay, there's time till then
1152[15:13:21] <Haohmaru> debian10 btw, is pretty nice, it just f*cking works \o/
1153[15:13:40] *** Quits: dvs (~hibbard@replaced-ip) (Remote host closed the connection)
1154[15:14:01] <Haohmaru> it looks more polished than 9
1169[15:20:39] <jelly> oxek, all software has bugs, at the very least opening a tcp service will allow for a possible DoS unless it has limits in place
1170[15:21:11] <jelly> what if someone's botnet opens a million connections to your nyancat port
1171[15:21:17] <Haohmaru> i'm scared to websearch that server
1172[15:21:25] <ratrace> hmm, I can't find the surveys page for Bullseye artwork...
1173[15:21:33] <ratrace> the link in mailinglist announcement doesn't work, replaced-url
1174[15:21:53] <Haohmaru> the default wallpapers and stuff get voted for?
1210[15:35:12] <Haohmaru> it's because things like this ^ that i feel scared when i'm websearching debian stuff and only getting results from askubuntu
1211[15:35:50] <ratrace> sudo which, while dirty, is quickest wya to check bin/sbin pathing issues :)
1213[15:36:02] <greycat> That's going to happen a lot, so you need to live with it. Just double-check anything you read against common sense and the manuals.
1214[15:36:05] <ratrace> guilty of abusing that, myself
1234[15:41:02] <oxek> talking of paths, I've successfully removed /bin and /sbin from /etc/profile on my system that has merged /usr
1235[15:41:25] <oxek> I probably gained a few % speed in looking up commands
1236[15:41:54] <jelly> that's not how hashes work!
1237[15:42:09] <greycat> I'm trying to imagine the Dwarf Fortress interface for system administration, and how it compared to the DOOM interface for killing processes.
1238[15:42:14] <oxek> hashes work only once the commands have been executed, no?
1277[15:57:59] <greycat> It also mentions "USB flash drive" up higher, so at least that hardware did exist at the time some of the documentation was written. I guess you'll need to dig deeper.
1278[15:58:37] <Eryn_1983_FL> hmmm
1279[15:58:43] <Eryn_1983_FL> ill just do with debian live disk and DD
1280[15:58:49] <Eryn_1983_FL> i dont need like NSA grade
1281[15:59:30] <hwm4rgs> i've ran dban from usb before, don't remember if i dd'd it or had rufus do its thing
1282[15:59:49] <Eryn_1983_FL> well ill try rufus
1283[15:59:57] <Eryn_1983_FL> is it nix?
1284[16:00:56] <ratrace> Eryn_1983_FL: dd once will make your data near invisible even to NSA. "military-grade" wipe recommendations are to make wipes FutureProof(tm). ie. who knows what discoveries lie ahead
1297[16:02:34] <ratrace> take your drives to a welding machine. nothing realigns ferromagnetic stuffs better than 90 amps driven through the chassis and plates :)
1298[16:02:35] <Eryn_1983_FL> but these i want to sell since they still good 2tb drives
1299[16:02:48] <Eryn_1983_FL> :)
1300[16:02:54] <jelly> ratrace, why is that better? metal platter disks typically do not optimize out written zeroes
1303[16:03:45] <ratrace> jelly: residual imperfections may give rise to detectable patterns if you apply zeroes uniformly. random data messes that up completely
1304[16:04:00] <ratrace> imperfections of magnetizing bits
1307[16:04:33] <ratrace> think of it like afterimage, aftre you close your eyes. now appply random noise to that.
1308[16:04:38] <jelly> if you have a state-level adversary or someone willing to spend 5-7 figures to detect those patterns you probably need a bit better opsec than asking about best practice on irc
1309[16:05:14] <jelly> normal $2000 data recovery won't find anything after even a single pass of zeroes
1311[16:05:18] <ratrace> how 'bout russian or chinese government scanning recycled/sold/thrown drives from western users and companies, in search of secrets?
1312[16:05:30] <jelly> see above.
1313[16:05:34] <Eryn_1983_FL> meh if they spend that much money on little of me they got bigger problems..
1314[16:05:41] <greycat> yeah, the chinese government wants all your pirated MP3s.
1315[16:05:41] <ratrace> jelly: you don' thave to be a specific target for that
1320[16:06:12] <ratrace> my server farm's ssh privkey is a secret I don't want made available because I threw out my drives.
1321[16:06:26] <jelly> again: if your data is that important, you need better opsec than asking about it in #d
1322[16:06:42] <Eryn_1983_FL> easy peasy crush it with your car ratrace
1323[16:06:52] <ratrace> they don't want my data, they don't know how I am. but if they had my ssh privkey? would they erase it because I'm a nobody to them?
1338[16:09:51] <ratrace> names, addresses, emails, password hashes (and the yprobably reuse passwords), phone numbers, some financial transaction records, etc.... that would be available if my server farm ssh privkey leaked.
1339[16:10:04] *** Quits: certaindestiny (5f615a5d@replaced-ip) (Remote host closed the connection)
1448[17:16:31] <SerajewelKS> running into a fun issue with btrfs + systemd. related to replaced-url
1449[17:17:42] <SerajewelKS> this means if i have a UUID= line in the fstab for a btrfs filesystem, and the /dev/disk/by-uuid/... symlink points to a device that _isn't_ the lowest device id in the btrfs volume, systemd will not auto-mount it. even worse, if manually mounted, systemd will automatically unmount it.
1451[17:18:04] <SerajewelKS> listing a specific device instead of a UUID has the same issue -- if the device with the lowest ID is replaced later, systemd will immediately unmount the fs when replacement is complete.
1452[17:18:13] <SerajewelKS> what's the workaround for this?
1456[17:24:02] <jelly> how did fedora solve that, if still true
1457[17:24:30] <SerajewelKS> no idea. systemd seems to lazy-unmount too.
1458[17:24:40] <jelly> I heard f32 defaulted to btrfs
1459[17:24:47] <SerajewelKS> which might even succeed for / and then the whole system is in a state where / can't be remounted until reboot
1460[17:25:22] <jelly> well one does not tend to remount / that often. anyway
1461[17:25:56] <SerajewelKS> doesn't matter... if your / is on a multi-device btrfs volume and you replace the device with the lowest device id, systemd will lazy-unmount /
1462[17:26:02] <SerajewelKS> i'm not sure what implications that has but they can't be good
1465[17:27:27] <SerajewelKS> the problem seems to stem from the fact that udev and /dev/disk/by-uuid assume that uuids are actually unique at the device level, which isn't true for btrfs
1466[17:27:34] <cluonbeam> jelly: f32 still defaults to xfs, afaik
1467[17:27:42] <SerajewelKS> only one physical disk can "own" a UUID in the udev world
1468[17:27:44] <cluonbeam> ext4 for /boot, xfs for everything else.
1471[17:29:07] <SerajewelKS> the solution seems to be to comment out the line in fstab and "systemctl daemon-reload" before doing any kind of reshaping of btrfs
1472[17:29:34] <SerajewelKS> then making sure that /dev/disk/by-uuid points at the lowest btrfs device before putting it back into fstab
1539[18:14:20] <spigot> looking for a vnc server that shares an existing X server, found tigervnc-xorg-extension. but dpkg-query -S on that package shows all it installed was in /usr/share/doc?
1562[18:31:06] <cluonbeam> I don't think I'll ever do a btrfs install.
1563[18:31:44] <ratrace> I running btrfs primarily to test it out. what bettre way than daily use? I don't yet dare to put in on the $$-making servers :) not _yet_ at least. maybe with upgrade to bullseye
1570[18:32:53] <ratrace> I do regular backups to a ZFS mirror
1571[18:32:53] *** cdown_ is now known as cdown
1572[18:33:20] <cluonbeam> As do I. I don't trust btrfs enough to not be a pain in my butt and have to restore.
1573[18:33:28] <ratrace> I'm surprised how well and stable it is. Last several times I tried it few years ago, it was complete mess.
1574[18:34:26] <ratrace> personally I'd love to see more support for dm-integrity. dm-integrity (with or without LUKS) under mdadm achieves same kind of bitrot protection and allows you to use any fs atop of mdadm.
1575[18:34:35] <cluonbeam> Suppose I can set my "get off my lawn" attitude aside and check it out.
1576[18:35:13] <ratrace> cluonbeam: that's what I did. I used to be a zfs fanboi and zealot. then zfs ate my data, then I flipped some tables and decided to put btrfs under proper testing.
1577[18:35:28] <cluonbeam> I've never had zfs eat my data.
1584[18:36:58] <cluonbeam> Is there a "right way" to do swap files on btrfs?
1585[18:37:10] <ratrace> cluonbeam: the license and current stance of kernel devs against zfs is kinda pushing me to love linux-native filesystems more. so far btrfs does it, even though it's inferior to zfs in some regards.
1586[18:37:27] <cluonbeam> ratrace: Oh, I agree with the stance against zfs.
1587[18:37:43] <cluonbeam> Anything Oracle has its fingers in is to be avoided like the plague.
1588[18:38:03] <cluonbeam> I don't generally have "religious convictions" about any given company, but Oracle is one of the few that I consider actively evil.
1589[18:38:11] <ratrace> same here
1590[18:38:49] <cluonbeam> Found my answer. Swap files are verboten prior to 5.0.
1602[18:45:18] <dpkg> perlmonkey: grow a sense of humor
1603[18:45:19] <ratrace> oh just one thing... my usage of btrfs depends on having a mirror. I wouldn't run btrfs on single drives as it's very, very fragile. simple corruptions can make whole filesystem unmountable without redundancy to heal itself.
1613[18:49:33] <cluonbeam> Okay, so, first btrfs question. Is it possible to turn a directory into a subvol, or do I need to treat it like a traditional mount point (evacuate the data, subvol creation, copy back)?
1622[18:54:42] <emilsp> is there a nice way to search for a package by the command/binary that package would install? I'm trying to figure out what I need to install to get modprobe.
1631[18:57:11] <ratrace> something tells me you turned root with su and didn't use -l so /sbin/ is not in your PATH
1632[18:57:14] <ratrace> !buster su
1633[18:57:14] <dpkg> In buster, su no longer overrides PATH by default, requiring that you use "su -" or "su -l" for login shells (which is not really a new thing at all...). To approximate the previous behaviour, put "ALWAYS_SET_PATH yes" in /etc/login.defs. See replaced-url
1671[19:06:27] <SerajewelKS> maybe i'm missing something but my understanding was that mv when both subvolumes are visible within the same mount (e.g. when a parent subvolume is mounted) is not cp+rm
1672[19:06:28] <cluonbeam> Yeah, I got that part, too.
1673[19:06:31] <ratrace> have to treat that part as any other mountable filesystem. copying data between filesystems.
1674[19:06:53] <SerajewelKS> if you mount the subvolumes separately then yes it would be cp+rm
1675[19:07:02] <ratrace> whether you cp + rm explicitly, or mv (which does cp+rm implicitly, across fs boundaries), is I guess irrelevant
1676[19:07:03] <SerajewelKS> but you could temporarily mount the parent somewhere
1677[19:07:30] <SerajewelKS> that's what i'm saying though, i don't think it's across an fs boundary when it's all done from within a single mount of a mutual ancestor subvolume
1678[19:07:45] <SerajewelKS> i was looking into this the other day and that's what i recall finding
1679[19:07:46] <ratrace> it is. subvols are distinct inode spaces
1685[19:08:21] *** Quits: julienth37 (~julienth3@replaced-ip) (Remote host closed the connection)
1686[19:09:23] <SerajewelKS> interesting. so if they are mounted separately i'm pretty sure it will do a full cp+rm but if within a single mount i would hope btrfs is smart enough to at least do a reflink copy?
1703[19:17:15] <ratrace> btw yea, one can cp --reflink between subvols too
1704[19:18:04] <ratrace> SerajewelKS: oh I know what you mean... yea btrfs does implicit mount on _that_ kind of subvols... uh not sure of the terminology here
1706[19:18:46] <ratrace> I don't think it'd reflink without explicitly being told to do so. I heard that future releases of cp would --reflink=always by default where possible, so there's that
1722[19:32:46] <doubletwist> So I'm looking to roughly match spacewalk's channel lifecycle capability with a local private debian mirror - such that packages/patches are added to dev at a certain point, then promoted to test when ready, then promoted to prod when ready.
1724[19:33:38] <alexrelis[m]> Is it possible to turn a standard LUKS encrypted partition into an LVM partition?
1725[19:33:39] <doubletwist> I will NOT use Katello (I really tried. I've given up). Is there any other solution other than storing multiple copies of the debian repos and just updating each in turn ?
1726[19:33:54] <alexrelis[m]> Or would I have to copy my entire partition back and fourth?
1748[19:55:01] <SerajewelKS> ratrace: like if you have subvolumes /one and /two, you can mount the parent subvolume / and then both of the subvolumes under look like directories and you could mv between them. i'd _expect_ btrfs to convert this into a reflink copy + rm but i don't know for sure.
1806[20:24:35] <greycat> The goal is for you to figure out what's using that partition. If you can find it, you don't have to have *us* find it.
1807[20:24:50] <cluonbeam> blebz: If a partition is in use, even partitioning the drive won't alleviate the issue, because the new partition table won't go live if the disk is in use.
1808[20:24:55] *** Quits: bocaneri (sauvin@replaced-ip) (Remote host closed the connection)
1895[21:16:03] <sney> blebz: why do you want to stop it? I thought you were trying to set up a software raid mirror. md127, with member devices sdb1 and sdc1, is a software raid mirror. you succeeded!
1896[21:16:31] <blebz> Well I had a raid setup earlier.. for some reason it seemed broken..the size of the 'disk' seemed small as I couldn't copy much stuff to it in the DE. It's the first time I've ever set a raid up on Linux.. so I thought maybe I botched something somewhere.. So I thought I would start again from scratch and now it's giving me this error..?
1897[21:16:46] <jelly> if your md is numbered 127 it was probably assembled automatically which you probably do not want to happen
1898[21:16:56] <sney> well you would need to destroy the mirror first if you want to create a new one from the same disks
1899[21:17:17] <greycat> when you say "start again from scratch", did you reinstall Debian without wiping out the RAID members on the disk?
1900[21:17:20] <brutex> which chan is for cloaks
1901[21:17:26] <jelly> I -think- some udev or systemd component can do that when it sees unknown md members appear
1902[21:17:26] <cluonbeam> #freenode
1903[21:17:34] <blebz> No I didn't re-install the server
1904[21:17:35] <cluonbeam> brutex: PS, cloaks do very little, honestly.
1905[21:17:48] <blebz> I probably could have by now..
1906[21:17:55] <sney> jelly: I think it's the kernel, I've seen md127 as the autodetect array for years
1926[21:25:51] <jelly> but then you get to support proprietary OSes that need a proper block device to be able to do certain kinds of backups, and they support iscsi but not nbd
1940[21:39:15] <ax562> Sup y'all? Is there a tutorial on editing grub.cfg after Debian Buster install to add a new operating system to choose at boot. I already have a multiboot system that works, I'm trying to add an lfs partition to the boot options menu.
1941[21:39:59] *** Quits: bigLITTLE (~saputra@replaced-ip) (Quit: This is so lame, I mean, why do you even need a quit message?)
1942[21:40:06] <sney> ax562: grub.cfg in debian is auto-generated by update-grub. you'll want to make your changes in /etc/grub.d/. there's a readme.
1956[21:44:00] <sney> any changes that you made to grub.cfg will be overwritten the next time you run update-grub, as the comments at the top of grub.cfg tell you explicitly
1957[21:44:06] <ax562> does it matter if my grub is on /sdx1 (first partition on the hd)?
1958[21:44:22] <sney> that's non-standard, but if your system boots then it's fine
1959[21:44:37] <ax562> sney update-grub will still work?
1960[21:44:48] <sney> yes.
1961[21:44:58] <ax562> ok cool. I will try! ty!
1962[21:45:22] *** Quits: Jerrynicki (~niklas@replaced-ip) (Remote host closed the connection)
1969[21:46:46] <greycat> No, you may edit /etc/default/grub to change some of the template values that update-grub puts into the generated file.
1970[21:47:07] <sney> you can't add boot entries from /etc/default/grub. both /etc/default/grub and /etc/grub.d are parsed by update-grub and can be used to customize the contents and behavior of the grub menu.
1974[21:48:14] *** dominic35 is now known as dominic34
1975[21:48:21] <jhutchins> Yeah, thanks. I've never actually had to add an entry with the current grub.
1976[21:49:20] <ax562> another question. How can I edit the order of the grub boot menu? On ubuntu I would just use grub-customizer but I do not think that worked or I couldn't install.
1977[21:49:56] <ax562> ooops didn't mean to ping another
1980[21:50:55] <greycat> if someone names themself "another" then they get what they deserve
1981[21:51:09] <ax562> lol
1982[21:51:39] <sney> the files in /etc/default/grub are numbered and that's the order that they go in the menu, iirc. and you can change which option is selected by default in /etc/default/grub.
1983[21:51:59] <greycat> no, you're thinknig of /etc/grub.d/
1984[21:52:00] <sney> it's a little finicky because of how debian organizes the menu by default
1985[21:52:07] <sney> right, mis-typed
1986[21:52:21] <sney> serves me right for trying to brain 2 monitors at once :P
1987[21:52:54] <sney> /etc/default/grub is not a directory; the numbered files are in /etc/grub.d
1994[22:01:19] *** Quits: gelignite (~gelignite@replaced-ip) (Quit: Stay safe! Stay at home! Stop the chain reaction!)
1995[22:02:00] <wouter> jelly: that's not true, actually
1996[22:02:19] <wouter> jelly: FreeBSD supports NBD (with a third-party translation between NBD and geom gate)
1997[22:02:28] <wouter> jelly: the Hurd supports (a very old version of) NBD
1998[22:02:32] <wouter> jelly: QEMU supports NBD
1999[22:02:59] <wouter> jelly: and someone is writing a Windows virtual storport driver for NBD, because Ceph uses it and they want to add Windows support
2025[22:25:38] *** Quits: thiras (~thiras@replaced-ip) (Remote host closed the connection)
2026[22:25:54] <foul_owl_> What is a good alarm clock program for debian? There used to be one I used for jessie but after upgrading sadly it seems to be gone
2066[22:47:51] <blebz> Actually that command: mdadm --stop /dev/md* seemed to completely break the box.. (I guess it could of been the uninstall of mdadm?) In any case the server rebooted into emergency mode... I got kinda freaked and I've now re-installed the box..
2067[22:48:23] <sney> !fundamentals
2068[22:48:24] <dpkg> You'll get more out of Debian and *nix if you learn some of the fundamentals. Read the "Unix and Internet Fundamentals" HOWTO at replaced-url
2069[22:48:54] <greycat> Breaking a server that you're able to reinstall is definitely going to be a learning experience.
2070[22:49:20] <greycat> Not one I would normally recommend, but... it sounds like it wasn't a total disaster.
2078[22:55:33] <blebz> Haha I'm still getting: mdadm: cannot open /dev/sdb1: Device or resource busy!!
2079[22:56:11] <blebz> Nooooooooooo oo o
2080[22:56:16] <greycat> Previously, I asked you whether you had reinstalled Debian but left the MD devices on the hard drive. You said no. But it seems now you have done exactly that.
2081[22:56:35] <greycat> Now MD is discovering the MD devices that are left over on the disk and automatically building an array of them.
2085[22:57:23] <blebz> It's in a magazine or I would ..
2086[22:57:41] <sney> print media. funky.
2087[22:58:13] *** rabbipires is now known as rowbee
2088[22:58:19] <sney> well, try getting better at using 'show' or 'list' arguments to commands, rather than just doing stuff and freaking out that it doesn't work.
2089[22:58:30] <sney> mdadm is great at giving you information when you ask.
2093[22:59:37] <sney> and as you already learned today, sometimes md arrays assemble themselves. now that you learned that, you can expect it, and respond to it
2098[23:01:02] <sverzel> Weird question. I have a perl script serving a web application with lighttpd. The script runs a system command and it's taking 3 solid seconds to return. Is it possible that the account used (replaced-url
2099[23:01:25] <sverzel> (running under my normal user, firing off the same system command takes 0 seconds)
2100[23:01:39] <sney> sverzel: I can't think of anything being throttled, but you can test it with 'sudo -u replaced-url
2101[23:02:07] <sverzel> sney, it's a nologin account, can I still sudo into it?
2102[23:02:43] <sney> you can run a command with sudo, you just can't get a login shell
2103[23:02:57] <sverzel> sney, ah okay, I shall try
2114[23:06:02] <sverzel> sney, the command is to use links2 to read a localhost site and dump out the formatted text. The web application performs other operations fast. Maybe I can try 127.0.0.1 just to eliminate..
2115[23:07:13] <sverzel> sney, nope, still slow :-) debugging continues..
2116[23:07:55] <sney> using a whole browser to pipe text? why not curl or something
2119[23:10:28] *** Tyszka is now known as JacobBlake
2120[23:10:35] <sverzel> sney, I'm after the text content of a HTML page. I know I can do it in more ways (also using Perl+libs). But not sure if curl can do that.. reading up on it though.
2133[23:23:12] <blebz> So if I do dd if=/dev/zero of=/dev/sdb (well both drives actually) .. then surely I gotta be able to start again and get this raid up?
2151[23:38:10] <robobox2> 1995 for me, heh. Odd I never learned about it in 25 years.
2152[23:39:57] <sney> hm, TIL
2153[23:39:59] <SerajewelKS> probably because it's only occasionally useful. usually if you're putting something else there anyway, you'll just mkfs or whatever.
2154[23:40:00] *** Quits: tagomago (~tagomago@replaced-ip) (Remote host closed the connection)
2156[23:40:13] <SerajewelKS> it's rare that you want to remove all magic signatures and then _not_ do anything with the drive for a bit
2157[23:40:28] <sney> indeed.
2158[23:40:40] <sney> I will probably forget again within the day, tbh
2159[23:41:05] <SerajewelKS> i've started using LUKS standard with everything so that when i sell/trash a drive i can just wipe the header instead of the whole device. wipefs will be useful there.
2168[23:44:40] <blebz> But I'm still getting: mdadm: cannot open /dev/sdb1: Device or resource busy
2169[23:44:41] <doubletwist> Ok I'm confused. why is debmirror only downloading source tarballs?
2170[23:44:42] <SerajewelKS> i think it requests exclusive access to the device
2171[23:44:57] <SerajewelKS> so it's probably mounted/used by something else in the system. lsblk could give you a clue.
2172[23:45:23] <sney> it's probably just the autodetected md127 again that keeps being a surprise to blebz despite dealing with it multiple times today
2173[23:46:10] <SerajewelKS> yeah md-raid likes to auto-assemble stuff any time there is a topology change anywhere in the system, even if it's not on a related device
2174[23:46:30] <SerajewelKS> e.g. writing a new partition table to a different device
2175[23:46:49] <doubletwist> using this conf: replaced-url
2176[23:46:59] <doubletwist> only doing contrib and non-free right now to test it
2189[23:54:51] <cluonbeam> blebz: You still haven't stopped whatever process is keeping the disk hostage. You're not going to fix the problem by changing anything on the disk.