22[00:07:15] <judd> [8086:24fd] is 'Wireless 8265 / 8275' from 'Intel Corporation' with kernel modules 'iwlwifi', 'snd-hda-intel', 'ata-generic' in stretch. See also replaced-url
23[00:07:58] *** Quits: DarkShymia (~dmenor@replaced-ip) (Remote host closed the connection)
24[00:08:05] <LtL> neilthereildeil: you have the proper module, but you most likely need the firmware, check dmesg for missing firmware.
55[00:15:46] <neilthereildeil> ok, and my understanding is that Microsoft has an agreement with Intel to distribute their firmware, but free Linux does not?
56[00:15:53] <LtL> neilthereildeil: add non-free and contrib to EACH line in sources.list then update and install it.
57[00:16:21] <LtL> neilthereildeil: pretty much yeah.
58[00:16:38] <LtL> everyone panders to microsoft
59[00:16:48] <neilthereildeil> LtL: oops, all i did was add this line: deb replaced-url
60[00:17:03] <neilthereildeil> and i installed iwlwifi
61[00:17:18] <neilthereildeil> but ur saying to remove that now and just add those 2 words to each EXISTING line?
62[00:17:19] <LtL> that'll work, now 'apt update' and install it.
63[00:17:22] <neilthereildeil> ok
64[00:17:41] <LtL> !sources.list
65[00:17:41] <dpkg> A suitable /etc/apt/sources.list for "Stretch" has three lines: "deb replaced-url
66[00:18:23] <LtL> neilthereildeil: i personally avoid httpredirect, use the factoid.
90[00:27:52] *** Quits: karakedi (~eAC53C340@replaced-ip) (Remote host closed the connection)
91[00:29:10] <rant> well ethernet firmware isn't special... but wifi companies apparently thing things like their noise floor calibration routines and such are.. so they make it closed source.. hence the issue
92[00:29:30] <rant> if they didnt ship blackbox firmwares, there would be no issue
97[00:30:25] <neilthereildeil> so the antenna and physics code is secret
98[00:30:28] <neilthereildeil> ?
99[00:30:39] <rant> they like to think so I guess :P
100[00:30:49] <neilthereildeil> heh interesting
101[00:31:09] <neilthereildeil> ok thanks for the help!
102[00:31:10] <whislock> It's not secret, per se, given that they're just implementing a standard. It's how they're implementing the standard in code that they consider proprietary.
104[00:31:48] <rant> maybe someday the mfgs and providers of service will realize what consumers actually want and are willing to pay good money for.. and its not being bullied and treated like criminals
161[00:59:19] <HeXiLeD> so i have this person that keeps opening windows of the browser, minimizes them and instead of maximizing them again, opens a new one (always for the same page). So i am thinking how to restrict the number of processes allowed for that app. What should I be looking for ?
162[00:59:51] <SerajewelKS> what problem is it causing?
163[01:00:00] *** Quits: dastier (~dastier@replaced-ip) (Remote host closed the connection)
172[01:01:52] <SerajewelKS> one process per app is going to break so much stuff
173[01:01:57] <HeXiLeD> even put a recording with instructions to play every 30 minutes
174[01:02:10] <rant> the browser won't even work with only one process
175[01:02:16] <HeXiLeD> rant: it is senior person that lives alone and uses the computer to talk to family
176[01:02:34] <rant> then make it more like a senior's computer.. less functionality
177[01:02:44] <HeXiLeD> that is what i have been doing
178[01:02:47] <SerajewelKS> are they opening another browser?
179[01:02:53] <SerajewelKS> or opening a new tab or something?
180[01:02:54] <HeXiLeD> everything is custom
181[01:03:09] <HeXiLeD> always a new window not tab
182[01:03:22] <SerajewelKS> maybe have the "start web browser" icon first check if the browser is running, and just bring that window to the foreground
183[01:03:31] <SerajewelKS> if that's how they're doing it
184[01:03:50] <rant> yes.. or.. as I already said, take the minimize function away :P
185[01:03:50] <Sveta> HeXiLeD, just leave the user with it - if they want to keep opening new windows and running out of memory it is their choice
213[01:06:41] <friendofafriend> HeXiLeD: Many will open the tab with that page already open, when you go to the same site twice. Some do not.
214[01:07:00] <friendofafriend> HeXiLeD: You may check out a similar query here. replaced-url
215[01:07:02] <SerajewelKS> your best bet is to customize the window decorations to remove all minimize buttons, or use a dock that minimizes to the same button as used to launch. then clicking the button just restores the prior window.
216[01:07:21] <mutante> HeXiLeD: reboot host once every 4 hours via cron ?:P
217[01:07:31] <HeXiLeD> SerajewelKS: a dock is even better
218[01:07:57] <SerajewelKS> i have to get going. good luck.
219[01:08:22] *** Quits: shpx (~shpx@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
220[01:08:34] *** Quits: Haudegen (~quassel@replaced-ip) (Remote host closed the connection)
228[01:13:20] <Ede|Popede> HeXiLeD: would an icon area help that's always visible? also i've seen a minimalistic wm before written in python without any window decoration. or a tiling wm. depending on the usage (only a few programs maybe) a click area with some wmctrl handlers or so may help.
233[01:14:36] *** Quits: lll3N1GmAlll__ (~lll3N1GmA@replaced-ip) (Remote host closed the connection)
234[01:14:47] <HeXiLeD> well the idea of a dock that minimizes to the same button as used to launch then clicking the button just restores the prior windo is actually perfect
249[01:18:13] <somiaj> hmm, I could see how to do that with FvwmButtons (sort of), basically make a dock button that launches the application using a conditional. If the application is running, just switch to it (un iconfiy it), other wise launch it if it isn't running.
250[01:18:53] <somiaj> I wonder if you could do this with other docs, so make a launcher use some program (maybe xprop) to check if the application is running and focus it if it is running, other wise run it. Then when you minizie it, just make it disapear.
251[01:18:56] <HeXiLeD> i am using idesktop for icons
257[01:23:52] <somiaj> HeXiLeD: you might be able to configure what apps turn into icons, and set up some launcher as I described. I can't say how to do this in different window managers or icon managers other than fvwm.
263[01:27:50] <somiaj> HeXiLeD: the wiki has some info, and learning to get info from the man page it is there. Though takes time, I don't know if there is a page for this exact situation. I think there might be info about the conditional launcher somewhere, one second.
265[01:29:07] <somiaj> as I said, it is a button that will do varous actions depending on conditions, like is the program running, is it active, etc. It will even close it (if you wanted)
266[01:29:29] *** Quits: jfoy (~jfoy@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
290[01:49:57] <GumShoe> Continuing my raid 1 experimenting from yesterday .... I have a pair of 4TB drives. I did a default minimal install on /dev/sda and am now trying to duplicate it in on /dev/sdb and a md. I used sfdisk to copy the part tbl from sda to sdb
292[01:50:28] *** Quits: a_l_b (~a_l_b@replaced-ip) (Remote host closed the connection)
293[01:50:49] <GumShoe> I try mdadm -v --assemble md126 /dev/sda2 /dev/sdb2 and get /dev/sda2 has no superblock - assembly aborted but I have booted from /dev/sda and /dev/sda2 is mounted
372[02:27:01] <SerajewelKS> in particular, lsblk will show you every disk, partition, and dm node (which includes encrypted volumes, dmraid volumes, and LVM volumes)
373[02:27:01] <coltkirk> i like how it lsblk shows mount points
374[02:27:20] <SerajewelKS> yep, that's pretty nifty too
376[02:28:17] <SerajewelKS> for awhile i was trying to figure out how to determine which physical partition owned which opened luks volume. cryptsetup doesn't seem to have a facility to list currently-opened volumes.
377[02:28:28] <SerajewelKS> you can dig around in /proc or /sys (i forget which) or you can just ask lsblk
382[02:29:39] <SerajewelKS> you have to be careful with lsblk though because some entries may be duplicated. for example, if you have LVM-in-raid1 (or any raid level with redundancy) you'll see the LVM volumes multiple times, under each physical device that provides the RAID.
383[02:30:05] <SerajewelKS> i guess you'll even see the RAID device multiple times
391[02:32:21] <SerajewelKS> you can very easily set it all up in the installer
392[02:32:39] <SerajewelKS> putting it in place after the fact isn't terribly hard either, if you know what you're doing (hint: i've set it up dozens of times, so it's not much of a challenge anymore)
393[02:32:50] <agio> but having to mentally map/reconstruct you set it up is challenging for me
394[02:33:02] <SerajewelKS> ah. it's second nature for me at this point.
395[02:33:11] <agio> especially when you add more layers (e.g. encryption ) over that
396[02:33:15] <SerajewelKS> i'd rather deal with RAID and LVM than physical partitions
413[02:37:26] <SerajewelKS> GumShoe: my disks are in a bit of unconventional RAID so i'll spare you the details, but basically all the disks have a 512MB dmraid volume which is the RAID1 /boot
414[02:37:52] <GumShoe> Are your drives larger than 2tb?
415[02:37:55] <SerajewelKS> GumShoe: the rest of the disk is also a RAID1. inside of the RAID1 is an LVM PV. the LVM VG has volumes for / and swap.
420[02:38:35] <agio> I guess the udev automatially assigned names (eg. /dev/disk/by-id/* etc) arent' available at the initrd stage eithher, becuase systemd hasn't run yet?
421[02:38:53] <SerajewelKS> i specifically put swap inside of the RAID1 so that the system can survive a disk failure without dying. it may seem smart to have two swaps, one on each drive. but if a disk dies and there is something swapped out onto that disk, the kernel is likely going to panic when it tries to swap that page back in.
422[02:39:21] <SerajewelKS> agio: the initrd has enough to probe for LVM and dmraid, so i would assume it also knows device names
423[02:39:26] <GumShoe> Don't you need a GPD boot? FWIW i've don't raid in centos/fedora but this is my first debian raid expereince...
429[02:40:16] <SerajewelKS> oh. yes, my disklabels are GPT. there is also BIOS boot partition for grub since my system uses BIOS boot and not UEFI.
430[02:40:23] *** Quits: Corshine7 (~Corshine7@replaced-ip) (Remote host closed the connection)
431[02:40:28] <SerajewelKS> grub understands GPT and dmraid RAID1
432[02:41:04] <SerajewelKS> i've had strange issues when i didn't have a separate /boot. grub was trying to boot from /-in-LVM-in-RAID1 and that usually worked but randomly wouldn't.
433[02:41:36] <SerajewelKS> i'm guessing the 3TB volume was somehow a bit much for grub, and if the kernel/initrd was allocated beyond _some point_ in the volume that grub couldn't load it. but i have no data to verify that assumption.
434[02:41:49] <SerajewelKS> switching to a dedicated 512MB mirrored /boot at the start of the disk solved that problem, though.
436[02:42:33] <SerajewelKS> but everything except /boot is in LVM-in-RAID1
437[02:42:51] <agio> setup dm-crypt physical volumes via debian installers patitioner, the kernel/initrd will automatically load the mappers for the system filesytsm and swap at boot. then mount the dm-crypt devices at boot
438[02:43:13] <SerajewelKS> agio: another huge advantage of LVM is that, if your motherboard can do SATA hot-swap, and you are able to physically install another disk without unplugging anything vital, you can use LVM to live-migrate a _mounted volume_ to another physical disk.
439[02:43:15] <agio> if you regularrly backup - no need for RAID
440[02:43:30] <SerajewelKS> RAID is not for backup in the first place
441[02:43:40] <SerajewelKS> whether or not you backup says nothing about whether you need RAID
444[02:44:18] <agio> isn't RAID primarily used for redundancy? i.e. disk failure?
445[02:44:26] <SerajewelKS> yes, but it's not about data loss
446[02:44:36] <SerajewelKS> you should always have backups in addition to RAID, if the data is valuable
447[02:44:43] <SerajewelKS> RAID is about uptime, not preventing data loss
448[02:44:58] <agio> ah, its a 'high availability' thing
449[02:45:14] <GumShoe> is/isn't /boot limited to < 2tb? And an additonal partitiong know on centos as biosboot is needed to?
450[02:45:17] <SerajewelKS> RAID is so a disk failure doesn't take down the server, and usually you can replace the disk with the server on as well (and the mirror will rebuild using spare iops)
451[02:45:48] <SerajewelKS> there are cases where you might want RAID but not need backups -- maybe it's a server where you _process_ a lot of data (and need it locally on disk so it's faster) but if the whole thing dies, you didn't lose any data
452[02:45:59] <agio> right
453[02:46:02] <SerajewelKS> but you can't tolerate the server going down due to a disk failure because you need it processing data 24/7
454[02:46:29] <SerajewelKS> likewise, you might want backups but not need RAID because you want to keep your data safe but can tolerate some hours/days of downtime. RAID and backups are orthogonal concepts.
455[02:46:48] <agio> can't you achieve that with load balanncing and database replication/sharding?
456[02:47:03] <SerajewelKS> (having said that, RAID can reduce the number of times you need to get your backups, but you should still have them)
457[02:47:11] <SerajewelKS> no
458[02:47:31] <SerajewelKS> that only protects you from the hard drives dying. that doesn't protect you from the sysadmin accidentally running a query that deletes everything
459[02:47:41] <agio> heh, yeah
460[02:48:27] <agio> someone here in IRC did that 2 nights ago, rm -rf'ed their $HOME
461[02:49:30] <agio> I've worked on a project where the mysql DB was hacked and the tables deleted too
467[02:50:24] <SerajewelKS> backups should be taken regularly, and stored somewhere that the system being backed up doesn't have alter/delete capability. it should only be able to add new backup data.
468[02:50:41] <SerajewelKS> if the system can alter/delete then you have no real protection from malware / accidents.
469[02:50:41] <agio> yeah,backup should be offsite
470[02:51:00] <SerajewelKS> well, yes, but you can do offsite backups where the system can delete stuff
471[02:51:13] <SerajewelKS> "offsite append-only" is the ideal
472[02:52:22] <agio> not sure if you remember our discussion few weeks ago about SSH forwarding?
473[02:52:50] <agio> I was wondering how you would forward _all_ ssh connections?
474[02:52:52] <SerajewelKS> ah, yes. i don't remember nicks that well, so i didn't remember that was you.
475[02:53:05] <SerajewelKS> what does "forward all ssh connections" mean?
476[02:53:24] <agio> well , I think we discussed this command
477[02:53:28] <karlpinc> Someone was talking the other day about the 3-2-1 rule. 3 backups. On 2 different kinds of media. With at least 1 copy offsite.
481[02:54:09] <SerajewelKS> karlpinc: indeed. and services like S3 and B2 are so cheap that it's silly not to pay a few bucks a month to sync a copy there, too.
483[02:54:23] <agio> which forwards a single connection
484[02:54:36] <SerajewelKS> okay. so what does "all" mean.
485[02:54:38] <karlpinc> SerajewelKS: So long as it's encrypted on the local end. :-)
486[02:55:09] <SerajewelKS> karlpinc: yes, we use restic which does local encryption and deduplication using a git-like content-addressable store, so that chunks of content including directories are stored by their hash.
487[02:55:16] <agio> "all" means, you open a SOCKS5 tunnel locally
490[02:55:28] <SerajewelKS> which also encrypts all metadata
491[02:55:34] <agio> right
492[02:55:52] <agio> how do you make all ssh connections go over that?
493[02:55:55] <SerajewelKS> agio: -D is useful but the client has to understand SOCKS, or you have to use something like socat
494[02:56:13] <karlpinc> SerajewelKS: My issue with restic is that it's entirely a push model. (IIRC) This does no good when the local box is compromised because then it has access and is able to destroy the backups.
501[02:57:02] <SerajewelKS> agio: connects to finalhost by tunneling through proxyhost, in one command
502[02:57:04] *** Quits: Ooze (~Ooze@replaced-ip) (Remote host closed the connection)
503[02:57:09] <agio> oh? thats simple too
504[02:57:17] <karlpinc> SerajewelKS: Nice.
505[02:57:32] <SerajewelKS> karlpinc: see replaced-url
506[02:57:48] <karlpinc> SerajewelKS: Append-only in the remote side REST config?
507[02:58:03] <SerajewelKS> karlpinc: note that rclone has a "restic serve" command which can work against any backend that rclone supports, and it understands --append-only. so you can run an rclone service that stores content directly in S3/B2/whatever.
508[02:58:07] *** Quits: Nekojimi (~jimj316@replaced-ip) (Remote host closed the connection)
509[02:58:17] <SerajewelKS> karlpinc: yes. so the server backups are pushed to will not accept any deletion/replace operations.
510[02:58:40] <SerajewelKS> except on "lock" objects, which is required to coordinate operations like prune/check
511[02:58:51] <karlpinc> SerajewelKS: I've been doing rsync hardlinked backups to save space. But I should take a look.
512[03:00:23] <SerajewelKS> karlpinc: i use restic to backup to a local disk and then i have a remote pull the repository with rclone copy, using --immutable
514[03:00:55] <SerajewelKS> karlpinc: which means that rclone will not delete anything, and it will not transfer any files that were changed. it only copies new files. and in fact, it fails if it detects a modified file.
515[03:01:08] <agio> SerajewelKS: say, I wan't to git clone something where I need the underlying ssh client (which git clone calls into) to use the ssh -J option
516[03:01:20] <SerajewelKS> this way i get speed when restoring from a local backup (don't have to pull over a WAN) but my offsite backups are secure because deletions/changes are not accepted by the offsite system
517[03:01:35] <SerajewelKS> agio: this is where your .ssh/config is handy
518[03:01:36] <agio> do I need to define the option in ~/.ssh/config?
519[03:01:47] <agio> under Host * ?
520[03:01:53] <SerajewelKS> agio: yes. the option you want is ProxyJump
522[03:02:22] <SerajewelKS> then you should be able to git clone from "foo"
523[03:03:32] <agio> is there a way to specify ProxyJump on a per command or per shell basis? e.g. can you export and environment var like `SSH_OPTIONS=ProxyJump' or similar?
524[03:04:17] *** Quits: deadz0 (~deadz0@replaced-ip) (Remote host closed the connection)
525[03:04:18] <karlpinc> agio: What are you trying to accomplish with that?
528[03:05:27] <SerajewelKS> agio: git has GIT_SSH_COMMAND, if that answers your question. i'm not aware of any facility ssh has.
529[03:05:37] <SerajewelKS> the real answer is to use .ssh/config
530[03:05:42] <karlpinc> agio: You can set up an "alternate hostname" in your ~/.ssh/config and use that hostname only when you want to do your git stuff.
531[03:05:49] <SerajewelKS> ^ was just going to say that
535[03:07:27] *** Quits: Sveta (~sveta@replaced-ip) (Ping timeout: 246 seconds)
536[03:07:38] *** Sveta_ is now known as Sveta
537[03:08:07] <SerajewelKS> karlpinc: the only thing that kind of sucks about restic is that, due to its git-like model, there's no way to quickly determine what can be discarded when you delete a backup
539[03:08:34] <karlpinc> SerajewelKS: How many backups do you keep?
540[03:08:54] <SerajewelKS> so it has a "prune" operation that scans all pack headers to discover all objects (particularly to clean up garbage from an interrupted backup) and then it does a mark-and-sweep on those objects by using the snapshot trees as the GC roots.
541[03:09:11] <SerajewelKS> the rest of the objects are removed. any pack containing a deleted object is rewritten.
542[03:09:40] <SerajewelKS> karlpinc: i use the policy-based "forget" options. so on our production servers, we keep 60 daily backups, 8 weekly backups, 12 monthly backups, and 2 yearly backups
543[03:10:12] <karlpinc> SerajewelKS: I just keep a few hundred daily backups. Probably 400.
544[03:10:39] <SerajewelKS> our repo is somewhere around 160GB and contains TBs of logical data
545[03:10:49] <SerajewelKS> 5 or 6 TBs worth of backups
546[03:10:57] <karlpinc> SerajewelKS: It wasn't worth paring down.
547[03:11:10] <SerajewelKS> yeah that was my thought too. storage is cheap.
549[03:12:01] <SerajewelKS> we have disk alerts on the backup server so i don't really run the forget script very often. though we sync to S3 for offsite, so i will run the forget script semi-regularly so we're not paying for storage for backups we don't need to retain according to the policy.
550[03:12:18] *** Quits: krabador (~krabador@replaced-ip) (Remote host closed the connection)
551[03:12:29] *** Quits: dez (uid92154@replaced-ip) (Quit: Connection closed for inactivity)
554[03:16:06] <GumShoe> During a debian 9 cdrom minimal install I'd like to wipe the partition tables on /dev/sda and /dev/sdb. I did this with fdisk and wrote the changes. But when I go back to the installer it's still seeing the /dev/mdXX raid devices I made.
555[03:16:38] <GumShoe> With the tools that are on the install cdrom how can I elminate the raid too?
564[03:18:21] <SerajewelKS> otherwise, go to the shell and lsblk to see what devices you have, then use mdadm to stop them
565[03:19:10] <SerajewelKS> if you have other dm devices (LVM LVs, crypt disks) then you should stop those as well
566[03:19:21] <SerajewelKS> it's more bulletproof to just reboot the installer if that's practical
567[03:19:37] <blackbart> When I first got my usb wifi dongle, I installed several different drivers (firmware-ralink from non-free, as well as compiling a driver offered by the manufacturer, and IIRC I got atleast one more off github). How can I find out which one I am using, and possibly switch to a different one?
578[03:27:14] <SerajewelKS> GumShoe: you were asking about my setup. my disks are 6TB, 3TB, 3TB, and 1TB. the 1TB disk has an LVM PV directly and is used as a non-redundant work area. (everything on that disk does not need to be backed up.)
579[03:27:48] <SerajewelKS> GumShoe: the 3TB disks form a 6TB RAID0
580[03:27:57] <SerajewelKS> GumShoe: then that RAID0 plus the 6TB disk form a RAID1
603[03:39:45] <agio> karlpinc: sorry for late reply - I was interrupted. cool, I'll give GIT_SSH_COMMAND a look
604[03:40:10] <GumShoe> So as I'm manually partitioning this install I see a partion type of reserved bios boot area. What about the ?magic? boot area that supports larger that 2tb drives? Will that just be configured without my intervention?
605[03:41:04] <agio> are you BOOTING FROM UEFI/GPT or BIOS?
606[03:41:41] <GumShoe> A 4tb drive
607[03:42:12] <agio> yeah but what boot firmware?
608[03:42:24] <GumShoe> Pretty sure I left the BIOS config set as UEFI. Can I check while the installer is running?
616[03:45:45] <GumShoe> If I'm set for UEFI would I see additional partition table type? To support the boot. I want to boot off either of the two raid1 devices in case of single drive failure.
619[03:47:17] <SerajewelKS> GumShoe: i did that because i previously had the two 3TB drives in a RAID1 and wanted to expand to 6TB. so i could buy 2x 6TB drives and find some other use for the 3TB drives, or get 1x 6TB and do what i did.
620[03:47:22] *** Quits: Insanity_ (uid179350@replaced-ip) (Quit: Connection closed for inactivity)
623[03:48:18] <SerajewelKS> GumShoe: the cool part is that i was able to do the reshaping all online. the process was basically this: grow the 3TB RAID1 to 3 devices, add the 6TB partition and wait for the array to rebuild. fail the 2x 3TB devices from the array and reduce the number of devices to 2.
630[03:50:14] <SerajewelKS> GumShoe: then i set up the RAID0 across the 3TB disks, and add the RAID0 device to the RAID1 array. once again, wait for the array to rebuild. once this is done, i ask the array to expand to be as large as possible (use the size of the smallest device) which changes the array from 3TB to 6TB. then i just have to resize the filesystem, which can be done online.
631[03:50:21] <SerajewelKS> so from start to finish this all required no downtime to accomplish
727[05:10:55] <SerajewelKS> GumShoe: a separate raid1 /boot is probably a good idea. there's very little reason for the rest of the devices to be their own separate raid1 devices. you'd be better off with a single raid1 that contains an LVM PV, then create /, /home, /var, and swap as LVM LVs.
728[05:11:07] <SerajewelKS> GumShoe: are you booting using BIOS or UEFI?
729[05:11:34] <SerajewelKS> GumShoe: honestly there's also little reason for /, /home, and /var to be separate volumes at all
733[05:13:08] <SerajewelKS> 4GB for /boot is also waaaaaaay overkill
734[05:13:16] <agio> rwp: thanks, trying now, but its not working. any idea how to pass the -vvv option to ssh when you are using ssh via 'git clone' ?
735[05:13:42] *** Quits: littlekitty (uid234006@replaced-ip) (Quit: Connection closed for inactivity)
736[05:13:52] <agio> also, would I need to allow agent auth forwarding from the proxy machine?
737[05:14:01] *** debhelper sets mode: +l 1458
738[05:14:18] <rwp> agio, Hopefully you have full ssh access and can debug this using pure ssh. But I guess not?
739[05:14:38] <SerajewelKS> GumShoe: here is my recommendation. the first partition on both disks should be a 1MB BIOS boot partition, for grub to be able to boot from GPT. if you are booting using UEFI, the next partition on both disks should be a 256MB EFI system partition.
740[05:14:56] <SerajewelKS> GumShoe: then have a 512MB physical raid partition on both disks, in raid1, with /boot inside.
742[05:15:31] <SerajewelKS> GumShoe: then the remaining space is another physical raid partition, in raid1, with an LVM PV inside. then create a 16GB swap LV and the rest of the space in the VG can be for /.
743[05:15:47] <SerajewelKS> you may want to leave ~5% of the VG space free for snapshots, which can be very useful to get atomic backups
746[05:17:07] <agio> rwp: the problem is the final machine is bitbucket.org , whilst - yes - I do have ssh key for the repo, but I can't run regular ssh commands (such as ssh me@remote uptime) to test connection, because the remote is not a full ssh login
747[05:17:13] <GumShoe> Problem I'm having now is that I switched my BIOS to UEFI and then did the partitoning again. It keesp complaining that I have to root partition I have / mirrored I did create a GPT partion first on both volumes but didn't mirror it.
748[05:17:29] <SerajewelKS> GumShoe: what is a "GPT partition"
759[05:19:12] <SerajewelKS> but having an EFI system partition on both disks takes a bit more work since you can't mirror it, so you have to keep it in sync some other way. it's a pain.
760[05:19:28] <SerajewelKS> the bootable flag is pointless with GPT
761[05:19:35] <SerajewelKS> i didn't think GPT even supported that flag
762[05:19:43] <GumShoe> pain 4 sure but I mustr be able to boot either disk if the raid 1 degrades.
763[05:19:57] <SerajewelKS> i prefer BIOS boot because keeping the EFI system partition in sync sucks.
766[05:20:32] <SerajewelKS> hmm apparently you can mirror EFI system partition with dmraid metadata <1.x
767[05:20:56] <SerajewelKS> because 0.9 stores metadata at the _end_ of the partition, meaning the UEFI system won't see the metadata ahead of the FAT header
768[05:21:40] <rwp> SerajewelKS, I believe that is correct. And I believe whislock also does this too.
769[05:22:10] <SerajewelKS> i don't think the installer supports selecting the metadata version though, and defaults to 1.x
770[05:22:17] <SerajewelKS> so you'd have to set that up in the shell :/
771[05:22:29] <GumShoe> current partitioning. I just made a huge / to see if I could get past the no root error. replaced-url
772[05:22:31] <rwp> It would need to manually set up that way with mdadm 0.90 raid.
773[05:22:34] <SerajewelKS> once you set it up, the installer partitioner should notice it, if you can make it refresh
779[05:23:53] <SerajewelKS> though if that is the case, it would be nice if the error message indicated that
780[05:23:55] <GumShoe> hmmmm...
781[05:24:37] <GumShoe> I'll try it for a test, but need to boot both drives eaisly....
782[05:24:42] <SerajewelKS> if you make a partition of the same size but empty (for the second ESP) then after you run grub-install you could dd the partition to keep it in sync. hacky though.
788[05:25:16] <SerajewelKS> which, again, is why i don't like UEFI boot
789[05:26:00] <SerajewelKS> it would be nice if dmraid allowed you to store the metadata in a small partition. you could have one raid metadata partition and one EFI system partition on each disk.
809[05:29:14] <SerajewelKS> then give yourself a 512MB raid partition on sda, and a 512MB on sdb
810[05:29:31] *** Quits: derlg_ (uid337581@replaced-ip) (Quit: Connection closed for inactivity)
811[05:29:40] <SerajewelKS> then allocate the rest of both disks as raid partitions. let me know when you're done, and post another shot of the partitioner.
818[05:33:53] <SerajewelKS> that can work if you back out of the partitioner and go back in, so it notices that things changed
819[05:34:21] <SerajewelKS> you get those partitions created?
820[05:34:24] <GumShoe> I tried this earlier today but couldn't assemble the raid arrays as /dev/sdaXXX was busy. I think because the partitions wern't marked as raid
827[05:35:48] <SerajewelKS> you want to create a raid1 of (sda2,sdb2) and another raid1 of (sda3,sdb3)
828[05:36:02] <SerajewelKS> show me a screenshot after completing these steps
829[05:36:46] <GumShoe> I did create the array with 'missing /dev/sdbx' when trying to --add I got sdaxx busy also tried some --manage --assemble foo...
836[05:41:39] <SerajewelKS> the goal here is to use LVM for everything but /boot
837[05:43:17] *** Quits: dvs (~hibbard@replaced-ip) (Remote host closed the connection)
838[05:43:38] <GumShoe> I'll stay with lvm through the experiment, but I''m not comfortable with LVM. Might have to RTFM LVM tomorrow. This may be a home server but I like to learn best or better practices in case I need the knowledge out in the 'real world'
875[05:55:07] <t3st3r> SerajewelKS> well, speaking for myself I got to dislike all this alignment, reservations and other quite troublesome inconvenient things.
876[05:55:10] <SerajewelKS> you may have to do a little dance to get sda1 formatted vfat again... the raid metadata clobbered the signature but the installer may not think it needs to reformat anything
877[05:55:40] <SerajewelKS> t3st3r: most of it is still necessary, for the ESP and for /boot
878[05:55:58] <t3st3r> not with btrfs fortunately :)
879[05:56:13] <SerajewelKS> ? you can't use btrfs for the ESP.
880[05:56:50] <t3st3r> hmm you mean EFI partition? Uh yeah EFI suxx.
881[05:57:14] <SerajewelKS> and my experience is that grub does strange things on >2TB volumes
882[05:57:29] <SerajewelKS> i've had multiple boot failures that disappeared when i created a dedicated /boot before the 2TB boundary
883[05:59:16] <t3st3r> Maybe it system firmware that does strange things? Though it rather known troublesome boundary if I remember. So yes, it gets some point.
886[05:59:45] <SerajewelKS> t3st3r: i've had the problem on multiple physical machines and VMs; i don't think it's a firmware/hardware issue
887[06:00:21] <SerajewelKS> GumShoe: yes. now we have to shift some stuff to get the partitioner to reformat the ESP that was clobbered by RAID. mark sda1 as "do not use."
890[06:00:44] <SerajewelKS> GumShoe: then go back into the software RAID configuration, and back out without making any changes.
891[06:01:09] <SerajewelKS> GumShoe: then mark sda1 as the ESP again, and go back into the software RAID configuration. you should get the "we need to format" prompt and it should tell you it's going to format sda1 as FAT.
908[06:11:03] <GumShoe> ok noted the change to fat then changed back to esp replaced-url
909[06:11:16] <SerajewelKS> and it formatted sda1?
910[06:11:34] <GumShoe> didn't try yet
911[06:11:42] <SerajewelKS> ok
912[06:11:54] <SerajewelKS> so when you choose software raid configuration, does it tell you it will format sda1?
913[06:12:08] <GumShoe> so 'finish part.. and write changes' ?
914[06:12:31] <SerajewelKS> no, select "configure software raid" again. we need to verify that it _will_ reformat sda1 and this is a convenient way to force that to happen.
915[06:12:44] <SerajewelKS> it should give you the "we need to write our changes" speech and it should specifically note that sda1 will be formatted
919[06:14:55] <SerajewelKS> basically it thought that sda1 was already formatted and so it didn't need to be. but when you created the raid device over it, that clobbered the vfat filesystem, but the installer didn't know that. so it won't reformat it unless we make it think you're changing sda1.
920[06:15:26] <GumShoe> it formated again. Not sure we still want two raids replaced-url
921[06:15:27] <SerajewelKS> i'm kind of surprised it even let you do that
922[06:16:17] <SerajewelKS> okay looks good. now let's set up the raid contents. under the 511MB raid device, select the 511MB "partition" #1 and use as ext, and mount on /boot
923[06:16:22] <SerajewelKS> this is our mirrored /boot volume
924[06:16:39] <SerajewelKS> then select the 4TB raid "partition" and use as: physical volume for LVM
937[06:20:40] <SerajewelKS> when that's done, select "configure the logical volume manager" and you'll get the prompt that it will format things, accept it again.
938[06:21:05] <SerajewelKS> in there you want to create a volume group. name it something pertaining to the machine you're building ("homeserver" or somesuch) and add the 4TB RAID device as the only physical volume)
939[06:21:13] <SerajewelKS> then you can start creating logical volumes
948[06:24:18] <SerajewelKS> GumShoe: once that is done, create a 16GB logical volume called "swap" and a final logical volume called "root" using the remaining space (just press enter when it asks you how much space)
949[06:24:44] <SerajewelKS> GumShoe: then "finish" configuring the LVM and it will take you back to the partitioner, where we will tell it how to use those two logical volumes.
950[06:24:50] <SerajewelKS> once we have done that, we can proceed with the install
957[06:26:33] <SerajewelKS> nice, now just create your 16GB swap logical volume and your root logical volume
958[06:26:59] <SerajewelKS> GumShoe: among other things, one of the nice perks of LVM is that you can move a _mounted_ volume between physical disks
959[06:27:58] <SerajewelKS> if you ever want to extend the logical volumes, make sure that any storage you add to the volume group is redundant (raid1) storage, otherwise you might as well not be using raid1 at all
1005[06:43:15] <GumShoe> When I go to make this production... is that huge Raid partion good? And then just create /tmp /home / and /var as LVM logical volumes? How big to make tmp?
1010[06:45:38] <SerajewelKS> there's nothing wrong with the huge raid partition. unless you know why you need /tmp, /home, and /var to be separate, you don't. many of our servers (set up before i was hired) have / and /var, but they made /var too small and it's caused us nothing but problems.
1011[06:45:57] <SerajewelKS> if you want to create them separately, yes, they would be separate logical volumes. i would argue that you only need /boot and /.
1012[06:46:07] <SerajewelKS> on debian, /tmp is a tmpfs (in-memory filesystem) anyway
1015[06:46:40] <SerajewelKS> regardless, it doesn't need to be its own filesystem. unless you want it to be a tmpfs.
1016[06:49:43] <GumShoe> when I use the guided partitioner it makes a bunch of temfs... I'm interested in having a seperate /tmp for increased security like mounting no suid and noexec
1022[06:53:49] <SerajewelKS> what i've learned is that when i try to make separate partitions for everything, i always guess their sizes wrong
1023[06:54:31] <SerajewelKS> if you do that, i would leave a good 50-25% of the volume group unallocated. then if a volume turns out to be too small, you can just use lvextend to allocate some of the free space to it, and you can grow the ext volume while it's mounted.
1025[06:55:10] <SerajewelKS> the nice thing about LVM is that volumes don't have to occupy contiguous space, so this is a valid strategy with LVM, whereas it wouldn't be remotely useful when using raw partitions
1033[07:01:40] <SerajewelKS> GumShoe: you could also create swap and a btrfs volume in the LVM
1034[07:02:09] <SerajewelKS> since multiple btrfs volumes can live in the same btrfs storage pool, they will all share free space, but can be mounted with different settings
1074[07:25:41] <GumShoe> @SerajewelkS Check it out! booted. Raid is up UU and I installed grub on /dev/sdb. You were very generous with your time tonight. Much appreicated replaced-url
1081[07:29:58] <karlpinc> GumShoe: In the FYI catageory, people here prefer paste sites that are text based, not images. Unless there's something graphical that needs to be communicated. For one thing you can't cut-and-paste from images.
1082[07:30:20] <karlpinc> !paste
1083[07:30:20] <dpkg> Do not paste more than 2 lines to this channel. Instead, use for text: replaced-url
1084[07:30:42] <SerajewelKS> GumShoe: since we don't have raid for sda1, you need to copy it. so: umount /boot/efi && dd if=/dev/sda1 of=/dev/sdb1 && mount /boot/efi
1085[07:31:00] <SerajewelKS> GumShoe: you should also fdisk /dev/sdb and change the partition of type of sdb1 to match sda1. after this you should be able to boot from either disk.
1086[07:31:23] <SerajewelKS> karlpinc: for much of the conversation he was in the installer
1087[07:31:32] <GumShoe> I have fpaste and cpaste generally but don't know how to use them with the screens. Hmm actually there is the screenshot button I could have rcp'ed that to one of my other servers
1088[07:31:38] <SerajewelKS> also in graphical mode
1097[07:34:19] <SerajewelKS> karlpinc: the EFI system partition must be FAT. so you _can_ raid it, but you must use 0.9 metadata so the superblock is at the end instead of the beginning. otherwise the UEFI system won't detect it as valid FAT.
1098[07:34:22] <GumShoe> @SerajewelkS Did you see that I installed grub on /ddv/sdb ?
1125[07:40:20] <whislock> It really doesn't. As long as it has the correct partition type, the "bootable" status is implicit.
1126[07:40:49] <SerajewelKS> karlpinc: hmm i wonder if using raid for the ESP is even a good idea. i was doing some research on this, and the motherboard UEFI could write some data to the ESP, which would desync the array.
1127[07:40:53] <whislock> Specifically, any partition with the correct type GUID is assumed to be bootable.
1128[07:41:04] <SerajewelKS> if you use a UEFI boot manager that's not aware of the raid, you will desync it as well
1129[07:41:43] <SerajewelKS> perhaps a better idea would be to mount sdb1 at /boot/efi2 and rsync them when running grub-install
1130[07:41:53] <GumShoe> what's the debian fpaste called? apt-get install ?????
1136[07:45:27] <dpkg> A command-line tool to send data to a <pastebin>. To paste e.g. your sources.list do "aptitude install pastebinit; pastebinit /etc/apt/sources.list"; to paste the output of a program do e.g. "dmesg 2>&1 | pastebinit". See also <pastebinit config>, <nopaste>.
1137[07:47:45] <GumShoe> @SerajewelkS Sure if you want to spend the time I'd like to properly configured /dev/sdb for boot. Here is df -h and df /devv/sdb replaced-url
1141[07:49:27] <SerajewelKS> GumShoe: (1) you can put them in a specially-crafted raid1, which will keep all linux-side changes in sync. however, this can be dangerous in that the motherboard UEFI system may choose to write data to the ESP. if it does so, it will only write to one or the other, not both. this will desync the array and linux won't notice until sometime later when it will notice the corruption, panic, and
1146[07:50:00] <GumShoe> So maybe I can make this my production server. I assume I can shrink the huge / LVM logical volume. Then I could create some mounts more to my liking.... Rather than reinstall it all tomorrow.
1147[07:50:02] <SerajewelKS> (2) you can format sdb1 and mount it at /boot/efi2 or something, and rsync the contents of efi to efi2
1148[07:50:15] <SerajewelKS> which you would need to do any time you reinstall grub (which will likely be rarely, if ever)
1149[07:50:37] <SerajewelKS> GumShoe: you can shrink / but not while it is mounted. you could boot the installer CD and shrink it from the command line though.
1157[07:53:31] <SerajewelKS> this would shrink the volume to 1TB
1158[07:53:51] <SerajewelKS> the --resizefs is important; that will shrink the ext4 volume to 1TB before truncating the logical volume. if you don't do that you'll hose the filesystem.
1165[07:56:01] <karlpinc> Shrinking a lv really safest to do in 2 steps. First use something like resize2fs to shrink the fs, then use lvreduce to shrink the lv.
1166[07:56:03] <GumShoe> O.K. more to read / learn. But you've got me as an LVM fan now...
1167[07:56:43] <SerajewelKS> karlpinc: if lvm doesn't know how to shrink the filesystem, --resizefs will result in an error. it's safe to do.
1168[07:57:02] <SerajewelKS> if you don't specify --resizefs then lvm will give you a warning that you have to confirm to proceed (without --force)
1169[07:57:21] <karlpinc> Note also that copying filesystem pieces to different volumes is best done in single user mode.
1170[07:57:39] <SerajewelKS> --resizefs does all of the usual steps: fsck, resize2fs, truncate lvm, resize2fs
1172[07:58:06] <SerajewelKS> karlpinc: i'd even go so far as to do it from the installer
1173[07:58:14] <SerajewelKS> since we have downtime anyway, may as well do it outside of the system
1174[07:58:21] <karlpinc> SerajewelKS: Sure.
1175[07:58:33] <karlpinc> (rescue mode)
1176[07:58:43] <SerajewelKS> though doing it that way should be fine, too
1177[07:59:43] <SerajewelKS> whislock: i did some research on the bootable flag / partition type and learned something interesting. not only does the partition not have to be bootable, it doesn't even have to have the correct partition type.
1178[08:00:02] <SerajewelKS> whislock: most UEFI firmware will look for a valid FAT filesystem containing the correct files and use it, regardless of flags/type
1182[08:00:33] <SerajewelKS> so the recommendation for raid1 ESP is to set the type to "linux RAID" so that tools don't accidentally use it as a raw FAT and desync the array
1186[08:01:51] <whislock> In practice, UEFI implemenations look for the first FAT32 filesystem they can find with /EFI/BOOT/BOOTX64.EFI in it, or whatever target exists in efivars.
1204[08:07:16] <SerajewelKS> if you want to switch to BIOS boot you can without reinstalling the system. you just have to change it in the bios, boot into the installer, fdisk both disks and change the partition type to the "BIOS boot partition" type, then grub-install on both disks
1205[08:07:24] <SerajewelKS> then a reboot should put you back in your system
1227[08:19:18] <GumShoe> On my current server I nave nagios setup to monitor raid on my home server among other services at home and the 'net at large..
1264[08:33:48] <karlpinc> GumShoe: Following random instructions on the internet is not a path to long-term stability. Get help here first, where we will tell you to use backports, and if that does not work get help from the upstream authors. Getting help from strangers is a last resort.
1292[08:46:26] <daggs1> Greetings, is there a way to add a pre up script to /etc/network/interfaces in a way that if the script returns 0 the iface will be inited and else it will not?
1318[08:57:27] <daggs1> whislock: I have a debian install with two vms, it creates 3 vnics, on one for direct network between the router vm and the host and the other two are virtual switch and connection between vm2 and the virtual switch that is connected to router vm, problem is, sometimes the vnet get mixed up resulting in the host trying to request ip for the wrong vnet
1319[08:58:09] <whislock> Huh, interesting.
1320[08:58:33] <SerajewelKS> GumShoe: i have just a few minutes to help. if you are available during business hours EST tomorrow, that's when i'll be at work at my computer.
1451[10:37:29] <thms> I just done a server migration using rsync / dest. I modified fstab. When I start the VM, after starting the services, I get no login prompt. Where should I be looking ?
1476[10:53:39] <AquaL1te> hi. i'm using a different debian version (raspbian) which uses /etc/network/interfaces (not sure if vanilla debian versions use something else). whenever i reload my networking.service unit, it fails. no warning about why, it just fails. however, it does set the addresses. i know that debian also (used) to use this config, that's why i also ask it here in the hope a guru can see what i'm doing wrong:
1482[10:57:43] <jelly> AquaL1te: on debian, restarting that service does not bring down and up all the interfaces. You'd use "ifdown -a" and "ifup -a" for that.
1485[10:58:04] <jelly> but you're not on Debian and I have no idea if raspbian has changed things
1486[10:58:08] <jelly> !raspbian
1487[10:58:08] <dpkg> Raspbian is a distribution <based on Debian> made specifically for the <Raspberry Pi>. Raspbian is not Debian and it is not supported in #debian. Please use #raspbian on irc.freenode.net for support. replaced-url
1488[10:58:55] <Habbie> jelly, yes, raspbian has changed things in this area
1489[10:59:23] <Habbie> jelly, also it might make sense to add #raspberrypi to that bot info thing
1655[12:32:51] <oo_miguel> blackflow: I did not find any settings in the bios. (beside boot from lan or something, but this has nothing to to with WOL I supppose, turned it on anyways)
1656[12:33:10] *** Joins: conta (~Thunderbi@replaced-ip)
1657[12:33:18] <petn-randall> oo_miguel: I'd first check if it works from suspend.
1658[12:33:35] <oo_miguel> petn-randall: suspend via "systemctl suspend"
1678[12:38:39] <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.
1779[13:33:56] <linuxthefish> hey, how can I stop network manager changing the resolv.conf file? I connect to different networks each day, so configuring each network to manual DNS every time isn't practical
1780[13:34:12] <linuxthefish> a cron job to set the DNS to google DNS is all I can think of...
1798[13:41:32] <AquaL1te> is there an easy way to compare update transactions? for example in fedora you can do `dnf history info last` to see details of the last update. also dnf shows this info during the output of an update. but with debian i seem to only see that an update has occured, but nothing specific to check running versions based on that. is that true?
1805[13:45:48] <petn-randall> AquaL1te: What do you mean with "but nothing specific to check running versions based on that"?
1806[13:46:00] <apollo13> linuxthefish: you obviously need to restart nm or reboot after the change
1807[13:47:19] <petn-randall> AquaL1te: I don't quite understand what you're trying to achieve. If you're looking for a transaction log, there's one at /var/log/apt/history.log.
1816[13:54:50] <AquaL1te> petn-randall: let's say you upgrade application x to version 2.2, and it was 2.1. there is nothing in the output or logs that indicates this upgrade path.
1904[14:38:21] <diogenes_> jelly, yes but if it's not installed correctly it might throw some errors
1905[14:38:23] <jken> basically I have a bootable image that writes another image out to disk, resizes some partitions, and installs grub, and I want to add a healthcheck to ensure grub installed correctly before rebooting
1906[14:38:28] <jken> jelly, ill check it out, thanks
1986[15:08:15] *** Quits: noodlepie (~Phillip@replaced-ip) (Quit: Rebooting to Linux kernel 4.20.2-Gentoo!)
1987[15:08:56] *** Quits: kupi (uid212005@replaced-ip) (Quit: Connection closed for inactivity)
1988[15:09:04] <bs0d> I've built an ncurses program named HelloWorld. It runs as ./HelloWorld fine, but /bin/bash -e ./HelloWorld gives me HelloWorld: HelloWorld: cannot execute binary file
1989[15:09:25] <bs0d> I do not understand what is wrong. How can I run my program by supplying it as an argument to bash?
1990[15:09:40] <greycat> "Doc, it hurts when I bend my arm this way!"
2109[15:50:20] <aloo_shu> hey, I'm on a laptop with a bad screen, trying to use an android tablet with an Xserver app instead. Setting/exporting DISPLAY=<ipaddr>:<displayno>, then starting a desktop session works like a charm, but: the tablet is also becoming the input device. Would there be a similarly simple way to set an environment variable that tells Xorg to use local mouse & kbd ?
2110[15:51:16] *** Quits: Silvering (~textual@replaced-ip) (Remote host closed the connection)
2111[15:51:35] <jelly> aloo_shu: it's X server that determines what are its inputs and outputs, you have to configure that wherever X server is running
2123[15:56:16] <aloo_shu> jelly: that would make it a lot more complicated, needing to forward mouse and keyboard etc. . As I understand it, X is involved in both sides, that is, w/o X installed, the local linux wouldn't be able to connect to the remote Xserver in the fist place
2129[15:58:34] <jelly> X clients can connect to a (local or remote) X server. You do not have a local X server that can use local input or output devices
2132[15:59:31] <petn-randall> jelly: they just rejoined, you need to ↑ \n
2133[15:59:43] <jelly> aloo_shu: X clients can connect to a (local or remote) X server. You do not have a local X server that can use local input or output devices
2162[16:12:16] <aloo_shu> gotta reiterate : [14:56] <aloo_shu> jelly: that would make it a lot more complicated, needing to forward mouse and keyboard etc. . As I understand it, X is involved in both sides, that is, w/o X installed, the local linux wouldn't be able to connect to the remote Xserver in the fist place
2210[16:37:06] <project2501a> hey guys, i keep getting this message every time i install a new daemon or service: "systemd synchronizing state with sysv service script"
2211[16:37:17] <project2501a> what's the point of sysv compatibility? and how do i remove it?
2212[16:37:33] *** Quits: krabador (~krabador@replaced-ip) (Remote host closed the connection)
2215[16:39:06] <petn-randall> project2501a: The point is so you can switch to sysv init without breaking your system.
2216[16:39:18] <petn-randall> project2501a: Why do you want to remove it?
2217[16:39:57] <project2501a> ok, understood and i can see where that may be needed. but i don't think i will ever go back to plain init. how do i disable this message and/or the compatibility, please?
2218[16:40:07] <greycat> Well, for most people, the point is that some services might not have systemd unit files written for them, and so the init.d compatibility layer allows those services to function.
2219[16:40:15] *** Quits: creat (~creat@replaced-ip) (Ping timeout: 252 seconds)
2242[16:48:15] <jelly> project2501a: you're in luck! You have time to file a Priority: wishlist or minor bug report, and help fixing it yourself!
2243[16:48:21] <jelly> !reportbug
2244[16:48:22] <dpkg> reportbug is used to submit bugs to the Debian <BTS>. Install reportbug, then run reportbug. See replaced-url
2245[16:48:24] <jelly> !debian-next
2246[16:48:24] <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.
2294[17:05:39] <jhutchins_wk> SerajewelKS: It's my understanding that recent SSD drives just present plain storage and don't need trim management, they handle that internally.
2307[17:08:05] <jhutchins_wk> petn-randall: I have a feeling that this is like the persistence of the rumor that lilo can't handle large hard drives for more than a decade after the problem was solved, or the rumors about sendmail long after it was secured.
2308[17:08:12] <SerajewelKS> frostschutz: dmraid
2309[17:08:25] <frostschutz> jhutchins_wk, SSDs will do wear leveling even if you don't trim but if you want to tell the SSD about free space, trim is pretty much it
2310[17:08:27] <SerajewelKS> frostschutz: there is no such thing as mdraid. you may be thinking of mdadm, the tool used to control dmraid.
2315[17:10:09] <frostschutz> SerajewelKS, they are different
2316[17:10:44] <SerajewelKS> frostschutz: "mdraid" doesn't seem to be a used term on google except by people who mean "dmraid" unless i am missing something.
2348[17:25:58] <frostschutz> SerajewelKS, well in userspace you have 'mdadm' and then you have 'dmraid' as utilities. mdadm does not control dmraid, it's a separate device /dev/md* /proc/mdstat, not /dev/dm-x dmsetup ... so it's a bit different. there is a dm-raid1 driver that does not share code with md (sometimes also called dm mirror) and a dm-raid driver that does share code with md but that doesn't make it quite the same
2356[17:27:57] <frostschutz> anyhow, in Linux you'd usually go with mdadm, and that works fine with raid1 SSD + HDD. set the HDD to write mostly for best reading performance
2357[17:28:13] <shibboleth> ##linux invite-only?
2358[17:28:26] <M6HZ> Hello
2359[17:28:31] <petn-randall> hi M6HZ
2360[17:29:48] *** Quits: encod3 (~encod3@replaced-ip) (Remote host closed the connection)
2361[17:30:09] <M6HZ> I encountered a bug with the latest "stable" version of xorg with an uptodate debian. xorg will crash if I try to log to another TTY and then disconnect from this TTY (ctrl + d)
2362[17:30:32] <M6HZ> Am I the only one to who this problem happens ?
2363[17:31:37] <jhutchins_wk> M6HZ: What happens if you type exit instead of Ctrl-D?
2400[17:42:16] <Zgrokl> it's not better to just use plain lvm raid ?
2401[17:42:18] <jelly> Zgrokl: both md raid and lvm2 are old, but fully supported
2402[17:42:38] <M6HZ> Is anybody able to reproduce the issue I encounter with xorg by switching to a TTY login to it, and exiting (ctrl - d ) this TTY to get back to xorg ?
2404[17:43:18] <jelly> Zgrokl: if you figure out how to replace members and look at rebuild process and manage bad blocks with lvm's raid features, let me know
2405[17:43:26] *** Quits: chele (~chele@replaced-ip) (Remote host closed the connection)
2406[17:43:32] <Habbie> jelly, are bad blocks still a thing?
2407[17:43:41] <jelly> Habbie: do you have SATA disks?
2449[17:53:25] <Ede|Popede> dammit. where is this "log to file" in xterm's menu. my finger hurts from pressing the fucking wheel and can't find it
2450[17:53:46] <M6HZ> Is anybody able to reproduce the issue I encounter with xorg by switching to a TTY login to it, and exiting (ctrl - d ) this TTY to get back to xorg ?
2451[17:53:55] *** Quits: vdehors (~vdehors@replaced-ip) (Quit: No Ping reply in 180 seconds.)
2459[17:54:57] <greycat> Nothing you said was clear at all except that you're using X (don't know how you started it) and you're switching VTs at some point.
2460[17:55:02] <M6HZ> greycat: That's not my question, I'm asking : Are you able to reproduce the issue ?
2465[17:56:35] <Phizzy> Debian Stable is well stable!
2466[17:57:02] <M6HZ> Leads to xorg crashing: step to reproduce on an uptodate debian OS. 1 - startx 2 - ctrl - alt - 2 (new TTY) 3 - login this TTY 4 - leave this TTY (ctrl - 2) 5 - Is xorg still working ?
2467[17:57:08] <M6HZ> greycat: clear enough now ?
2468[17:58:07] <Zgrokl> also if i have 64gb of ram, is it good to have a swap of the same size ?
2469[17:58:11] <Zgrokl> or 8 gb will do ?
2470[17:58:13] <greycat> Tested on Debian 9. Ctrl-Alt-F2, login as wooledg, Ctrl-Alt-F1. I am still here.
2507[18:07:23] <greycat> If that shows any missing firmware filenames, you can try installing them.
2508[18:08:09] <aloo_shu> I might be re-asking, my connection is shit. Problem: I want o run a remote X session, but keep the mouse&kbd input on the machine the desktop session is running on, not the one with the remote display server. Any suggestions how to -tell the local Xorg my wishes, or -add the remote Xserver as an output that RandR can see?
2527[18:14:42] <judd> Package x2x (x11, optional) in stretch/amd64: Link two X displays together, simulating a multiheaded display. Version: 1.30-3; Size: 27.6k; Installed: 69k; Homepage: replaced-url
2528[18:15:13] <jelly> it's an old piece of software, I have no idea how well it works
2582[18:27:31] <dpkg> UNIX does *not* store the creation time of files. Anywhere. You see that "ctime" field in the stat structure? It's not "creation time". It's inode (metadata) change time. It gets updated when you chmod the file, etc. NEWSFLASH: ext4 stores creation times after all, ask me about <crtime>.
2585[18:27:51] <wwilliam> im not a good source to ask stuff
2586[18:27:59] <wwilliam> I am quite stupid and slow
2587[18:27:59] <xand> > sarge
2588[18:28:02] <xand> lol
2589[18:28:08] <SerajewelKS> !sarge
2590[18:28:09] <dpkg> Sarge is the codename for Debian GNU/Linux 3.1, released June 6th, 2005. Sarge security support ended on 2008-03-31, this release is no longer supported. Sarge users should upgrade to Etch, ask me about <sarge->etch>. Removed from the mirrors; ask me about <sarge sources.list>, <sarge-backports>. Get old sarge ISOs here: replaced-url
2591[18:28:16] <SerajewelKS> jeebus
2592[18:28:24] *** Quits: n4dir (~n4dir@replaced-ip) (Quit: Lost terminal)
2593[18:28:40] <AK> !mpd
2594[18:28:40] <dpkg> Music Player Daemon (mpd) is a server allowing remote access for audio file/stream playback and playlist management. Prior to version 0.17-1, Debian's mpd package did not use <dmix> by default (Debian bug #568588). replaced-url
2595[18:29:04] <jelly> wwilliam: you question as it is does not make a lot of sense, best ask whoever knows the details to come ask themselves
2596[18:29:20] <GumShoe> Goodmorning @SerajewelkS I'm just trying to find our conversation late last night in irc #debian logs...
2644[18:38:54] <jelly> wwilliam: if you have shell access you can use "find" command to figure out which files have a modification time newer than local midnight
2652[18:40:44] <wwilliam> nice just what i need it thanks jelly
2653[18:41:14] <jelly> !using find
2654[18:41:14] <dpkg> using find is probably replaced-url
2655[18:41:31] <xdruppi_> why do "recommended" packages exist? i mean aren't they kind of like suggested ones?, is it possible to configure apt to not install recommended by default?
2672[18:46:37] <dpkg> Packages required in all but unusual cases (that is, "breakable" dependencies) are called "recommended" packages. By not installing these packages, you miss the standard functionality they provide. To disable once, "aptitude -R install packagename" or "apt-get --no-install-recommends install packagename". To disable permanently, echo "APT::Install-Recommends no;" > /etc/apt/apt.conf.d/no-recommends. See <why recommends>.
2675[18:46:58] <dpkg> From Debian 5.0 "Lenny" onwards, apt-get and aptitude both install "Recommended" packages by default. From <policy> section 7.2, Recommends, "declares a strong, but not absolute, dependency. The Recommends field should list packages that would be found together with this one in all but unusual installations." You're not that unusual, trust us. By not installing "recommended" packages you will be missing functionality.
2712[18:54:52] <maxrazer> Has anyone experienced Firefox being slow compared to Chromium? A fellow debian user has told me he doesn't use Firefox for that reason.
2713[18:55:32] <AK> maxrazer: firefox is slower than google chrome. I didn't use chromium
2718[18:56:08] <Floflobel_> hello, I installed a new Debian server but the keyboard is not good, what is the method to change it? I just tested by following this wiki but it does not propose a keyboard selection. replaced-url
2729[18:58:42] <judd> [8086:0122] is '2nd Generation Core Processor Family Integrated Graphics Controller' from 'Intel Corporation' with kernel modules 'i915', 'snd-hda-intel', 'ata-generic' in stretch. See also replaced-url
2731[18:59:05] <maxrazer> I haven't been using Firefox, because the binary package is compiled to require pulse audio and I haven't gotten around to doing a custom build.
2732[18:59:10] <greycat> I've been googling but without a lot of success... it's a "Sandybridge" which is from 2011, so it may be old enough not to use non-free firmware.
2737[18:59:58] <jelly> dpkg, tell Floflobel_ about keymap
2738[19:01:09] *** Quits: m0j0dj0dj0 (~punk3r@replaced-ip) (Quit: go drink with my bitches!)
2739[19:01:14] <cybrNaut> maxrazer: a fundamental difference between Firefox and Chromium is Chromium spawns a separate process for every tab. So i think it's expected that each tab would render fast, but Chromium will also bring your whole system to a crawl
2747[19:03:11] <tomodachi> maxrazer: not really , just checked i have 50 tabs approx , only 4 processes
2748[19:03:54] <maxrazer> tomaw, I haven't looked into it much, but that link says you can adjust that, so maybe you can increase that in the settings.
2749[19:03:57] <Floflobel_> jelly, what about keymap ?
2753[19:04:17] <tomodachi> maxrazer: yup seems like it default is 4
2754[19:04:19] <cybrNaut> maxrazer: i believe mozilla claims Firefox has "sandboxing" for security purposes.. each tab in a sandbox of some kind, but it's still lesser of a sandbox than separate processes
2755[19:04:23] <jelly> Floflobel_: check your client for a private message from dpkg bot
2756[19:04:35] <jelly> different tab/window perhaps
2815[19:20:44] <jelly> Floflobel_: is there any difference if you run with C locale, "LC_ALL=C dpkg-reconfigure keyboard-configuration" ? Are you logged in at the console or via ssh?
2880[19:46:41] <uio> Hello! On LXDE here. Is there a way to have an on screen feedback when I use the volume keys to change the volume? Ie, a little icon would show up, down or off?
2881[19:46:41] *** Quits: beaver (~none@replaced-ip) (Remote host closed the connection)
2913[19:56:43] <diogenes_> it has nothing to do with debian
2914[19:56:45] <nirakara> even after i open them
2915[19:57:01] <nirakara> diogenes_: are you talking to me?
2916[19:57:08] <n_1-c_k> uio: if you find no easier way, you can use xbindkeys to set the volume and show a message by e.g. osd_cat. I don't know how you'd set a tray icon.
2940[20:05:09] *** Quits: kapil____ (uid36151@replaced-ip) (Quit: Connection closed for inactivity)
2941[20:06:29] <uio> Lubuntu on Ubuntu forums recommends apt-get install xfce4-volumed gstreamer0.10-als neither of which are available in the repos...
2953[20:10:43] <rwp> Tom01, It has always "just worked" for me. Left button down, drag to highlight, up to stop. Then middle button pastes the highlighted region.
2954[20:11:02] *** Quits: starch (~user@replaced-ip) (Quit: rcirc on GNU Emacs 27.0.50)
2966[20:12:56] <greycat> In X11, pasting is done with the mouse. You highlight with the 1st (left) button, and then you click the 2nd (middle) button once to paste the highlighted text.
2968[20:13:28] <greycat> Some APPLICATIONS (web browsers mostly) will also mimic the Microsoft Windows keyboard-based pasting paradigms, but that's on a program-by-program basis.
2973[20:15:11] <e> well, properly, the ctrl+c and/or ctrl+insert thing is a different selection from the middle mouse one, they don't replace each other
3001[20:21:41] <SerajewelKS> now we have to figure out redundancy for the EFI system partition
3002[20:21:45] <GumShoe> ... lurking ... Just started a re install to apply all I learned yesterday. LVM seems so much easier today....
3003[20:21:53] <rwp> But I think life is a lot more peaceful if people don't know what others are thinking.
3004[20:22:12] <SerajewelKS> LVM is fairly easy but of course you still have to learn it. once you know what's going on, it's dead simple and really nice to work with.
3009[20:23:20] <GumShoe> I had a lot of learning curve to learn the installer. I didn't find it all that intuitive.... One thing I really wish it had was a clone /dev/sda to /dev/sdb kind of feature....
3010[20:23:20] <SerajewelKS> my #1 desire for linux storage management is to have online shrinking of ext*
3012[20:23:57] <rwp> Years ago an experienced admin was setting up a system for me and used LVM. I asked, Why? It wasn't needed. He said he didn't have an answer beyond that he always uses it because it makes things easier. Years later *I* was setting up a system for someone and used LVM and they asked me, why LVM? I had that same look on my face then as my mentor had years before.
3013[20:24:19] <SerajewelKS> yeah it's hard to explain why to use it except that when you need it, you're glad it's there
3017[20:25:04] <SerajewelKS> make sure you take regular backups if you use btrfs. of course, you should take regular backups regardless of what fs you use.
3018[20:25:22] <rwp> SerajewelKS, I once did a shrink of a much to large ext3 of aroudn 600GB and it took 12 days to complete!
3019[20:25:25] <SerajewelKS> GumShoe: if you dive into btrfs you'll have to pick someone else's brain about it. i know about it conceptually but haven't actually used it.
3034[20:27:51] <SerajewelKS> rwp: interesting... btrfs supports online grow and shrink
3035[20:27:52] <greycat> You know what else nobody ever does? Decide that they need LESS space.
3036[20:27:59] <GumShoe> Uh ohh *fear* back to ext4 I do have regular backups on all my linux machines at home and the cloud even a ms server 2008R2. I use bacula, but am thinking of switching to some Rsync scripts. Are there any _popular_ rsync based script projects?
3039[20:28:42] <SerajewelKS> greycat: it can be useful in cases like adding encryption/raid to an existing volume that fills the disk. you usually have to shrink by an MB or two to accomodate the raid/luks superblocks.
3040[20:28:51] *** Quits: kesenai (~username@replaced-ip) (Remote host closed the connection)
3043[20:29:10] <SerajewelKS> i've also shrunk database server volumes so that we can make use of LVM snapshots for truly point-in-time backups with no downtime
3044[20:29:12] <jhutchins_wk> We use LVM so that we can grow a volume without taking it off-line. Add a disk in VMWare, expand the volume, no downtime.
3046[20:29:18] <rwp> I have been using BackupPC for some years and it is solid for a small number of systems, say a dozen, but gets very slow as systems are added to the collection.
3047[20:29:35] <rwp> I have been thinking of exploring other backup options...
3048[20:29:43] <SerajewelKS> rwp: my go-to these days is restic
3049[20:29:44] <jhutchins_wk> I've also worked on a system where the tiny overhead of LVM was enough to take the system down.
3063[20:32:33] <dpkg> "Does anyone have X or use Y?" is taking a poll, not asking a good question that IRC helpers can answer. Don't do it or sussudio's army of militant badgers will hurt you. Also see <ask> and <bad polls>.
3130[20:45:00] <AK> I use efistub(direct boot the kernel without an extra boot manager) and I need a hook which could rename the kernel and initrd on upgrade to vmlinuz and initrd.img respectively. I mean there should be no versio with them
3134[20:46:04] <uio> So I install volti, which seems to have volume indicator capacities, but I can't get them to work... how can I use this to get an on screen volume display?
3147[20:54:39] <GumShoe> Here I am in the installer at the partitioner using manual methond. Is there a way to wipe the drives from the gui? Or just grab a shell and wipe them which would be easier then deleting/removing all the lvm then raid stuff...
3148[20:54:45] *** Joins: conta (~Thunderbi@replaced-ip)
3149[20:55:28] <deepy> What's the driver for intel hd 4000 package called? I can only find xserver-xorg-video-intel and that seems to be the wrong one
3174[21:05:55] <GumShoe> The annonying part of that is reboot! ;-) Anway I'm back in the partitioner and it occurs to me that I just need to redo the LVM logical volumes. I also changed the /dev/sdb1 from part to /boot/efi to match /dev/sda1
3184[21:08:13] <judd> [8086:0166] is '3rd Gen Core processor Graphics Controller' from 'Intel Corporation' with kernel modules 'i915', 'snd-hda-intel', 'ata-generic' in stretch. See also replaced-url
3237[21:42:56] <Resilience> I need to uderstand the ouuput from a dig search, due to an issue with dnsbl, he says I am running some kind of mail proxy (but I cannot uderstand what he says in his webpage)
3238[21:43:34] *** Quits: guru (~guru@replaced-ip) (Quit: Leaving)
3254[21:47:02] <rant> Resilience: no its great judt the pronouns are wrong :p
3255[21:47:31] <rant> which is common of american latinos
3256[21:48:02] <Resilience> rant, no, they are not wrong, dnsbl is a one man effort, his name is Leandro whatever, and he thinks he is some kind of internet sheriff or vigilante
3261[21:49:19] <GumShoe> @SerajewelKS Finished the reinstalled with the LVM logical partitions created and left some extra space for growth. replaced-url
3262[21:49:21] <rant> that may be so but gramatically it is an it
3263[21:49:56] <GumShoe> Looking pretty good but didn't get the sdb1 created as /boot/efi
3265[21:50:48] <rant> English doesnt assign gender to things that dont reproduce sexually and if the lgbtq folk get their way it wont assign gender at all leavin us all confused
3272[21:52:08] <Resilience> rant, no, he is not an it, becuase dnsbl is run by one man, so a man it still a he, annd a woman still a she, ight? so the man hiding beneath dnsbl has done that, so it has been done by him
3273[21:52:25] <Resilience> dnsbl does nothing, it is all done by one man, so it is a he
3284[21:55:16] <SerajewelKS> GumShoe: you have to figure out how you want to tackle the ESP. use the 1.0 superblock and hope nothing desyncs the array, or manually rsync/dd after reinstalling grub (which you should never have to do).
3285[21:55:20] <jmcnaught> Resilience: try sharing a pastebin of the dig output, you might also try ##email for help with this issue
3293[21:57:58] <GumShoe> Last night I installed base utilities. Today I didn't could that explain the error; Failed to contact the server: [Errno socket error] [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed (_ssl.c:720)
3297[21:59:27] <GumShoe> I think I have my install to my liking so time to add packages after initial install. Is this a prefered way to find packages? tasksel --list-task
3298[21:59:38] <jmcnaught> Resilience: I don't use any dnsbl myself, but I believe you look up by reversing the octets in the IP address you want to check and appending the dnsbl domain name, for example 1.2.3.4 becomes 4.3.2.1.dnsbl.example.com
3299[22:00:09] <GumShoe> I was expecting to see some dev groupinstall and such...
3300[22:00:31] <SerajewelKS> GumShoe: wait something is wrong
3301[22:00:40] <SerajewelKS> GumShoe: why is md0 not mounted as /boot?
3302[22:00:46] <Resilience> jmcnaught, mmm the problem is that the man running dnsbl has black listed me, and some irc servers have banned me, and I don't understand what LEandro from dnsbl says in his webpage
3314[22:03:32] <jmcnaught> Resilience: you're not really checking anything with that dig query, and you got no result. Like I said I don't use dnsbl myself, but if you are trying to check if your IP is listed by this one, do they not have instructions for how to check?
3317[22:05:31] <GumShoe> I didn't modify the raid. I figured we had the raid correct last night so just redid the LVM. I did change try to change /dev/sdb1 to /boot/efi but it didn't take.
3319[22:05:48] <Resilience> jmcnaught, as I said, I don't understand what they say about rDNS and things like that, I don't know what he _thinks_ he has detected and I cannot check it, as I said Leandro Carlos rodrigues thinks he is some kind of vigilante of the inteenet and he asks for money to delist banned IPs from his service
3320[22:06:26] <GumShoe> @SerajewelKS Guess I'll wait to install additional packages until we fix md0
3339[22:11:42] *** Sonnenmann is now known as PaddyF
3340[22:11:56] <jmcnaught> Resilience: if you do "dig -x <IP address>" then you will get the reverse DNS that points to that IP address. With dig you can specify which nameserver to check using "dig @ns.example.com …" You might need to use both of those options to check the DNSBL, but I don't know anything about it in particular.
3342[22:13:18] <Resilience> jmcnaught, the man running that service list IP's, he says nothing of why they are listed, sys that if he dones not like waht you said will ban youur email too, and asks fo rmoney por paid delisting
3349[22:16:42] <Resilience> jmcnaught, that is the uqestion, I have NO DNS service, NO mail server, NO web server, he does not say what he thinks I am running, not how can I check it to understand what he thinks I do
3350[22:17:00] <Resilience> jmcnaught, so, rDNS of WHAT? of nothing?????
3351[22:17:17] * Old_Dog comes in but appears to be sleepwalking
3352[22:17:38] <jhutchins_wk> Resilience: What blacklist do you think you're on?
3353[22:17:57] *** Quits: banisterfiend (~textual@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
3356[22:19:37] <Resilience> jhutchins_wk, I don'think, I am at dnsbl, I have checked for it, and I don't understand why, so I was asking what can I do to understand why I am listed, becasuse the man at dnsbl says "configure our rDNS", but I do NOT have any service
3357[22:19:50] <SerajewelKS> GumShoe: if the partitions were the same, the installer won't use them unless you explicitly tell it to
3358[22:19:58] <SerajewelKS> GumShoe: because it assumes you are installing alongside of an existing system
3359[22:20:10] <SerajewelKS> GumShoe: so you still have to tell the installer that md0 is /boot or it will just leave it alone
3360[22:21:36] <GumShoe> What is /boot used for? Considering I have /boot/efi ?
3363[22:21:43] <jhutchins_wk> There is no "dnsbl". There are quite a few different ones. Mostly you get listed for spam coming from your address. Some of them have poor policies, which is why you should know their reputation before using them for blocking as opposed to weighting.
3365[22:22:41] <SerajewelKS> GumShoe: we can fix it like this: mkdir /boot2 && mount /dev/md0 /boot2 && rm -fr /boot2/{*,.*} && mv /boot/{*,.*} /boot2/ && umount /boot2 && mount /dev/md0 /boot
3366[22:22:45] <Resilience> jhtsorry, you're right, is this one: replaced-url
3367[22:22:47] <SerajewelKS> GumShoe: then we have to put /boot in your fstab
3368[22:23:03] <SerajewelKS> GumShoe: /boot/efi stores the grub image. /boot stores the kernel image, initrd, and grub config+resources
3378[22:26:25] <jhutchins_wk> Resilience: It's possible that your IP address was spoofed by a spammer and that got you listed. It's also possible you got malware and sent spam you didn't know about.
3395[22:30:12] <Resilience> jhutchins_wk, Leandro, from spfbl says soemthing about rDNS, but if I am not running anything DNSed cannot RDNS it, so I have to go to he paid delist, he is a blackmailer
3398[22:31:03] <Habbie> Resilience, let's start at the beginning - you want to connect to an irc server. the irc server does not let you in. what does it say?
3399[22:31:14] *** Quits: Levure (~quassel@replaced-ip) (Remote host closed the connection)
3404[22:32:33] <Resilience> Habbie, it says I am in some blacklist, so I go to a blak list lister, check for my ip and I am only listed in ONE site, the spfbl
3405[22:32:46] <Habbie> the black list lister may be incomplete
3406[22:32:50] <Habbie> you should talk to the irc server admin
3407[22:32:51] *** Quits: towo` (~towo@replaced-ip) (Remote host closed the connection)
3408[22:32:51] <greycat> Resilience: no, what is the *exact* text of the message that you see
3431[22:35:36] <Habbie> it's more likely almost the entire internet is banned
3432[22:35:38] <greycat> There's a form on this page where you can request to be removed from the blacklist (G-line). You'll need a rudimentary grasp of Spanish.
3433[22:35:39] <Habbie> with some exceptions
3434[22:35:40] <SerajewelKS> GumShoe: yes, the command line i gave you would migrate that into md0
3436[22:35:52] <SerajewelKS> GumShoe: when you installed you didn't tell it to use md0 for /boot, so /boot was just created as part of /
3437[22:36:10] <greycat> I would suggest that you fill in the form, using Google translate if you have no local Spanish-speaking friends who can help you.
3438[22:36:10] <Habbie> Resilience, i don't think this is related to any DNSBL
3439[22:36:17] <greycat> OK, excellent.
3440[22:36:24] <Habbie> greycat, i believe it was established earlier that ... that
3441[22:37:49] *** booyah is now known as booyah_
3442[22:38:05] <Resilience> greycat, , Habbie the page says I am at some bleck list, or running malware, as I am running debian, I assume I am NOT running malware, go to th webpage to check my listing, and the page says I am listed at spfbl
3444[22:38:37] <greycat> You are assuming things that make no sense. Just fill in the form and request the removal of the G-list, which is entirely within this IRC server and has nothing to do with email blacklists.
3452[22:38:52] <Habbie> Resilience, did replaced-url
3453[22:38:53] <SerajewelKS> Resilience: means your IP has been sending spam. either you do have some malware, or someone else at your residence has malware, or your IP was recently rotated and the last person who had it had some malware.
3459[22:39:20] <Habbie> SerajewelKS, in context, that seems to be irrelevant
3460[22:39:23] *** booyah_ is now known as booyah
3461[22:39:30] *** up`ime is now known as ravioli
3462[22:39:52] <greycat> He's banned by this IRC server, so he googled his IP and found that he is *also* on some email blacklist, and he has been following the steps on the email blacklist page instead of the steps on the IRC server's page.
3463[22:40:02] <Habbie> right
3464[22:40:25] <Resilience> greycat, Habbie do you understand Spanish? have you ever being hit bofere by this issue? I do, both of them, I understand what the page says, and contacted before the irc network, they said I have to be delisted, my ISP changed my IP's and the I could log in into the network again
3465[22:40:26] <Habbie> and the text on the irc server page is obviously wrong
3466[22:40:44] <Habbie> Resilience, i'm reading what google translate makes of the page now (after having, with some effort, read the Spanish)
3467[22:40:56] <Habbie> Resilience, it is quite obvious that the text on the page makes no sense
3468[22:41:01] <greycat> Yo hablo solamente un poquito.
3469[22:41:08] <Habbie> Resilience, so, did you, or did you not, fill out the form on replaced-url
3473[22:41:14] <GumShoe> @SerajewelKS I updated that pastebin with those commands to check syntax before executing. replaced-url
3474[22:41:22] <Resilience> Habbie, I am a native Spanish speaker, and can hardly understand what the page says
3475[22:41:53] <Habbie> Resilience, 'usually does not last more than 1 hour' 'infected with trojans' 'problems in your router' are the highlights in the translation
3476[22:41:53] *** CaptainN is now known as KevinKeene
3477[22:42:04] <Habbie> Resilience, none of those make sense to me
3479[22:42:19] <Resilience> greycat, , Habbie what I am REALLY asking is how can I use dig to know i is there soem kind of DNS pointing at my ip, a web server, a mail relay, or somwething like that
3480[22:42:28] <greycat> You can't.
3481[22:42:32] <GumShoe> FWIW Interesting I fixed that pastebin error by 'tasksel install standard'
3517[22:49:22] <Resilience> habiethe problem with spfbl is that hey don't say WHY some IP is listed, they just say "put the RDNS on", but, if I don't have any service, wht do I do?
3519[22:49:40] <Habbie> Resilience, the IPs i checked, that are on the spfbl, have fine rdns
3520[22:49:46] <Resilience> Habbie, the problem with spfbl is that they don't say WHY some IP is listed, they just say "put the RDNS on", but, if I don't have any service, wht do I do?
3528[22:50:44] <SerajewelKS> GumShoe: do "cd /boot" (because you are probably still in the old boot)
3529[22:51:03] <Resilience> Habbie, well, if you read to Leandro from spfbl, he says the solution from being FREEley delisted is having he rDNS configurated rig, bt if no, you go to the pay delisting
3536[22:52:23] <SerajewelKS> GumShoe: hmm no, the umount failed which means something else did too
3537[22:52:30] <SerajewelKS> GumShoe: ls -l /boot /boot2
3538[22:52:37] <Resilience> Habbie, maybe, but goig back to my origina queston, how can I check the is no entry to my IP? reverse DNS look up not checked for that?
3539[22:52:53] <Habbie> Resilience, reverse DNS is the only one you -can- check with dig
3572[22:59:22] <greycat> If your Spanish IRC server insists on an ISP-provided (worthless) PTR record, then you probably won't be able to connect to them directly from this ISP. Maybe you can rent a VPS or something and bounce your connections from there.
3573[22:59:23] <Habbie> Resilience, on whatever the reverse pointed to
3574[22:59:28] <Habbie> Resilience, but your reverse points nowhere
3575[22:59:41] <Habbie> greycat, that's not it - i tried the spanish server from two IPs that have fine reverses with matching forward
3576[22:59:53] <SerajewelKS> GumShoe: aha yes, efi. that's fine.
3577[23:00:09] <SerajewelKS> GumShoe: umount /boot/efi && umount /boot && mount /dev/md0 /boot && mount /boot/efi
3578[23:00:13] <SerajewelKS> wait
3579[23:00:20] <SerajewelKS> i'm dumb, one sec
3580[23:00:21] <greycat> Well, I don't see what #debian can do for him.
3581[23:00:33] <SerajewelKS> GumShoe: umount /boot/efi && rmdir /boot/efi && umount /boot2 && mount /dev/md0 /boot && mount /boot/efi
3587[23:01:44] <greycat> If they have a vendetta against your ISP, there's not much you can do unless you can bounce your connections from a different source.
3588[23:02:00] <Resilience> Habbie, greycat, I don't really understand what is a "reverse DNS" I thought that "forward DNS was URL into IP" and he reverse DNS was "IP into URL ponting to IP"
3594[23:02:39] <greycat> A reverse DNS is formally a PTR record, which is registered within the global Domain Name System and says "this IPv4 address maps to this fully-qualified domain name".
3595[23:02:46] <Resilience> Habbie, no URLS? how they are called them? server names?
3596[23:02:54] <Habbie> Resilience, domain names, sometimes host names
3598[23:03:15] <Resilience> Habbie, well, change url to HOSTNAME :)
3599[23:03:16] *** Quits: karakedi (~eAC53C340@replaced-ip) (Remote host closed the connection)
3600[23:03:28] <greycat> "dig +short -x 188.240.132.226" gives no output, so you have no PTR record
3601[23:03:38] <Habbie> don't use +short, you'll never know why it's not there
3602[23:04:02] <Resilience> Habbie, greycat, I don't really understand what is a "reverse DNS" I thought that "forward DNS was HOSTNAME to get IP" and he reverse DNS was "IP into HOSTNAME pointing to IP",was I wrong?
3603[23:04:09] <greycat> A reverse DNS is formally a PTR record, which is registered within the global Domain Name System and says "this IPv4 address maps to this fully-qualified domain name".
3604[23:04:33] <Habbie> Resilience, reverse is IP pointing to hostname.
3617[23:06:47] <johnkeates> well, it was just 3 lines above where reverse dns was thought to be a pointer while a pointer is a pointer and reverse dns is ambigious
3618[23:06:54] <SerajewelKS> GumShoe: perfect. make sure /boot/efi has stuff.
3619[23:06:57] <SerajewelKS> GumShoe: now you can rmdir /boot2
3620[23:07:10] <johnkeates> (reverse DNS would actually be a reverse lookup)
3663[23:14:39] <Resilience> greycat, I thing that the idiot running the extortion scheme from spfbl thinks I have an open ealy, maybe I have misconfigured my exim (maybe, not an expert)
3664[23:14:50] <Habbie> again
3665[23:14:55] <Habbie> spfbl makes no sense
3666[23:15:10] <greycat> I know that you have no email service, because you said it earlier. Other people seem to be trying to get you to set one up, which seems nonsensical.
3667[23:15:16] <GumShoe> @SerajewelKS Want a paste? I saved it in a file...
3668[23:15:22] <mutante> how about watching the exim logs and if it moves really fast you have an open relay
3669[23:16:06] <mutante> or just forget about running a mail server at home.. ack
3670[23:16:17] <SerajewelKS> GumShoe: sure, or just copy the UUID for /boot
3671[23:16:20] <GumShoe> @SerajewelKS the uuid is short xxxx-yyyy in sda and sdb
3672[23:16:22] <Resilience> greycat, ah sorry I get lost, I am not running a mail sever, but I am running exim for inernal mail, so maybe it is misconfigured, can I chekd some how if I am an open realy?
3673[23:16:26] <greycat> But since you're a home user on dynamic IP, you probably have a web space set up somewhere else, like a VPS or whatever. Use that to probe your home computer if the answers we've already spend a LARGE amount of time giving you are not satisfactory.
3674[23:16:33] <SerajewelKS> GumShoe: we don't care about that, we care about the UUID for md0
3675[23:16:39] *** Quits: war9407 (war@replaced-ip) (Remote host closed the connection)
3676[23:16:42] <Resilience> mutante I cannot forget what I never knew :)
3677[23:16:47] <mutante> Resilience: if you are not sure if it's misconfigured then it means you dont know if you are running a mailserver
3679[23:17:30] <mutante> if you did dpkg-reconfigure exim4 and selected just the "satellite" stuff then it's likely just fine though
3680[23:17:35] <SerajewelKS> GumShoe: ok. what does "mount" say about /boot? (what fs)
3681[23:17:41] <Resilience> greycat, no, you were not answering anything you (all) have passed more time tryig to deny me and argumeting that answering what I asked, but that is another question
3695[23:20:36] <SerajewelKS> GumShoe: okay so it's ext3. so open up /etc/fstab in your editor, and add this line before the line for /boot/efi: UUID=9d136b06-28b9-49b5-9796-dac5c96ac362 /boot ext3 defaults 0 0
3712[23:25:13] <Resilience> Habbie, my last question (just sorry for beiing so DNS ignorant) as my computer is not listed in any registered service AFAIK, it cannot have a rDNS, right?
3713[23:25:39] <Habbie> Resilience, i don't understand your question, sorry; but your IP does not have a reverse DNS PTR today, no
3714[23:25:40] <Resilience> mutante, nothing moves, only the "night moves"
3715[23:26:04] <greycat> AN rDNS (a PTR record) is done by your ISP. It has nothing to do with "services" that you are "registered" for.
3716[23:26:13] <GumShoe> can I comment out the line in /etc/default/grub GRUB_CMDLINE_LINUX_DEFAULT="quiet"
3717[23:26:24] <SerajewelKS> GumShoe: if you would like, sure
3718[23:26:44] <GumShoe> I like to have the verbose output...
3719[23:26:47] <greycat> Many ISPs just create useless boilerplate PTR records like customer-180-100-180-100.mydomain.com
3720[23:27:02] <greycat> Your ISP didn't.
3721[23:27:03] <ghormoon> hi, can I do something like paste -d";" aaa bbb that would actually put it like "line 1 of aaa;line 1 of bbb" ?this just pastes all lines of aaa (with ; appended) and all lines of bbb (with; appended)
3722[23:27:15] <Habbie> greycat, it's possible they did but nobody sees it because the delgation is missing, but yes
3723[23:27:54] <Resilience> Habbie, as I said, I don't understand completely the DNS, so I don't have the right vocabulary, but his issue has caught me, I was just asking if having not a registered domainname nor hostname, cannot get one, the problem doe snot come from there, right?
3724[23:28:05] <Resilience> his*this
3725[23:28:09] *** Quits: greycat (~wooledg@replaced-ip) (Quit: They see me clawin' the love seat / They won't do nothin' 'cause I'm cute and furry)
3726[23:28:14] <ghormoon> oh nevermind, i just have the input malformed :D
3727[23:28:19] <Habbie> Resilience, this is not about whatever domain you own - this is about your ISP, who owns the IP space, not doing reverse DNS
3728[23:28:59] <Resilience> Habbie, ok, but, usualy IPs use to have a reverse DNS if there are no DNS entry defined? just curoios
3729[23:29:10] <Habbie> question does not make sense
3737[23:31:40] *** Quits: mr_robot (~mr_robot@replaced-ip) (Remote host closed the connection)
3738[23:31:47] <Resilience> Habbie, oh man, try to understand, for an IP of a residential user as me, who has no DNS registered service at that IP, no web server, no mail server, no anything, just a a broadband router and a personal computer, is it usual to have a rDNS for that? if that, would be the first notice for me
3749[23:33:33] <SerajewelKS> GumShoe: okay run that first, then check /boot/efi
3750[23:33:41] <Resilience> Habbie, that was my question, I thougt it shouldn't but with this issue I thought I could be wrong
3751[23:33:46] <GumShoe> ls /boot/grub has some ifi stuff fonts grub.cfg grubenv locale unicode.pf2 x86_64-efi
3752[23:33:57] <mutante> Resilience: yes
3753[23:33:58] <mutante> "RFC 1912[1] (Section 2.1) recommends that "every Internet-reachable host should have a name" and that "for every IP address, there should be a matching PTR record," it is not an Internet Standard requirement"
3754[23:34:01] <Resilience> mutateshould they exist for every IP?
3755[23:34:06] *** Quits: Slashman (~Slash@replaced-ip) (Remote host closed the connection)
3756[23:34:16] <mutante> note the "Internet-reachable" part
3758[23:34:49] <Resilience> mutante, I think that part is erchable by the DNS, right? not by IP addrees, but by DNS name
3759[23:35:18] <storrgie> I'm seeing something quite confusing... I am on stretch and have two buster lxc containers, it looks like all the services are somehow in the "same cgroup" so I can see services from both containers listed in each one of them (e.g the container that doesn't have nginx installed can see nginx.service when I `systemctl status`, but there is no nginx process)
3761[23:35:40] <Habbie> mutante, also note that RFCs are not in fact allowed to make operational demands
3762[23:35:55] <mutante> Resilience: all of this is the DNS system. you can look up name to IP or IP to name .. hence it's called reverse when looking the other way around
3778[23:39:22] <mutante> Resilience: i never asked you stuff, i answered your question. i am not trying to be smarter than you but now i lost interest. cya
3780[23:39:50] <SerajewelKS> GumShoe: nice. okay so now, i would reboot and make sure everything comes back up.
3781[23:39:53] <Resilience> mutante, you're worng, you were asking to waht you wanted
3782[23:40:10] <SerajewelKS> GumShoe: and then you still have to decide if you want /boot/efi to be an array, or if you want to periodically rsync/dd it to keep it in sync
3784[23:40:39] <SerajewelKS> GumShoe: using an array is dangerous in that your motherboard UEFI implementation could alter the ESP on one disk but not the other, and then the array has desynced but linux doesn't know it
3785[23:40:50] <mutante> Resilience: where did i even form a question? are you mixing up people?
3790[23:42:28] <SerajewelKS> GumShoe: no, as long as you put /boot in there we should be good
3791[23:42:36] <Resilience> mutante have you lost interest? or do you want to argue with me for trying to be smrtaer than me as I sadi, you're smrtater than me, you know way betetr the DNS system, I wouldn't be hre asking OTHER things than that if not for another "smarter than the others" guy, who rans spfbl
3792[23:42:40] <SerajewelKS> GumShoe: we had to do that and then update-grub/grub-install so that grub knows where /boot is
3793[23:42:48] <SerajewelKS> GumShoe: we did all that, so it should boot fine now
3807[23:47:06] <SerajewelKS> i wish that grub-install was smart enough to locate and automatically mount the ESP on the disk it's installing to, instead of assuming /boot/efi is the place to go
3808[23:47:15] <SerajewelKS> if it did that, we wouldn't need hacks to make the ESP redundant
3809[23:47:52] <GumShoe> @SerajewelKS Submit a pull request! </ducks>
3810[23:48:07] <SerajewelKS> GumShoe: TBH i would just manually copy its contents
3811[23:48:15] <SerajewelKS> GumShoe: the contents will only change when you grub-install
3812[23:48:45] <GumShoe> @SerajewelKS Do I infer it's not worth mirroring it?
3817[23:49:41] <SerajewelKS> GumShoe: if you wind up booting from sdb because sda failed, then it might still fail to boot because /boot/efi in fstab doesn't have the nofail option, and mounts by UUID
3818[23:49:54] <SerajewelKS> though, hmm. after dd'ing they will have the same UUID. i wonder how the system will handle that.
3852[23:57:11] *** Quits: t3st3r (~t3st3r@replaced-ip) (Remote host closed the connection)
3853[23:57:12] <SerajewelKS> GumShoe: when you install a new kernel, a new kernel is placed in /boot and the config under /boot/grub is recreated. /boot/efi is not touched.
3854[23:57:26] <SerajewelKS> GumShoe: fdisk looks perfect. i think you're done.
3856[23:58:30] <GumShoe> @SerajewelKS Except testing.... I'll go try booting with each of the other disks unplugged. Then wait all night for the md rebuild...
3857[23:58:44] <SerajewelKS> GumShoe: the md rebuild should take seconds due to the write-intent bitmap