DocScrutinizer05 | <kou_2> Hey, so turns out that udev breaks if a btrfs balance is paused before rebooting - as in, it times out and waits for the balance to start up again before remounting so you get dropped into emergency shell | 11:32 |
---|---|---|
Wizzup | why would udev do that... | 11:34 |
KotCzarny | because, who knows, maybe some not-well-tested distro? | 11:34 |
DocScrutinizer05 | I been mentioned in devuan context | 11:36 |
KotCzarny | maybe btrfs refuses to mount before unpausing | 11:36 |
DocScrutinizer05 | but... why would udev _not_ do that? | 11:36 |
KotCzarny | disclaimer: never used btrfs | 11:36 |
DocScrutinizer05 | btrfs calance is a critical atomic procedure I'd guess | 11:38 |
DocScrutinizer05 | balance* | 11:38 |
DocScrutinizer05 | sth about btrees afaik | 11:38 |
DocScrutinizer05 | oops sorry, not devuan, suse context | 11:42 |
DocScrutinizer05 | so you'd assume the fs they have as root default for their enterprise product should be damn well tested, at least when compared to the average community driven linux | 11:44 |
Wizzup | I can't imagine why this would be an issue | 12:07 |
Wizzup | or why udev would test for this | 12:07 |
Wizzup | I've done this many times with btrfs and eudev anyway | 12:07 |
DocScrutinizer05 | I can imagine why it wou,kd be an issue, when the fs was in an inconsistent state so system can't mount rootfs | 12:36 |
DocScrutinizer05 | but I never tested any of this, so I don't support any notion on it | 12:37 |
KotCzarny | let's blame it on the cluelessness | 12:37 |
KotCzarny | and move on | 12:37 |
DocScrutinizer05 | possible | 12:37 |
Maxdamantus | The overall balance isn't atomic, but I think the reallocation of each block is. | 12:38 |
Maxdamantus | actually, I don't exactly "think" that, but it seems likely that it would be. | 12:39 |
Maxdamantus | the overall thing certainly isn't though, since it's just reallocating a bunch of chunks. | 12:39 |
Maxdamantus | (should have said "chunk" before instead of "block") | 12:39 |
Maxdamantus | note: a "balance" in btrfs involves rewriting a number of chunks storing the data, usually to either uniformly distribute the chunks across devices, or to change to different "raid" levels. | 12:42 |
Maxdamantus | and unless you apply a filter, it involves rewriting the entire filesystem. | 12:42 |
KotCzarny | cute little operation | 12:42 |
KotCzarny | even windoze handled defrag better | 12:43 |
Maxdamantus | It's a background thing. | 12:43 |
Maxdamantus | It shouldn't normally interrupt anything. | 12:43 |
Maxdamantus | dunno what that person's initramfs is doing. | 12:43 |
Maxdamantus | Things shouldn't normally care about the balance. Normally if you shut down or crash during a balance, it will just continue when you mount it again. | 12:44 |
Maxdamantus | but it's not critical that it continue .. you can always just cancel a balance if you want to. | 12:45 |
KotCzarny | for some reason it had to do something important and wasnt allowed to continue, which resulted in unmountable fs | 12:46 |
Maxdamantus | If it's unmountable at a kernel level, you should almost certainly get at least some feedback in the kernel log. | 12:46 |
Maxdamantus | It shouldn't be unmountable just because a balance was in progress/paused. | 12:47 |
KotCzarny | s/sunmountable/mountable/ | 12:47 |
KotCzarny | he might have some serious problem apart from that pause | 12:48 |
Maxdamantus | on a couple of occasions I've had to manually zero the log before being able to mount (which supposedly means potentially losing up to the last 30 seconds or so of writes), but I don't think I've done that in the last couple of years, so presumably they do that automatically now in the kernel, or they've just fixed some of the issues that led to an unusable log. | 12:51 |
Maxdamantus | afaik that used to be a typical reason for not being able to mount. | 12:51 |
Maxdamantus | also regarding defrag, balance doesn't help with fragmentation. | 12:54 |
KotCzarny | yeah, i was just joking on it | 12:54 |
Maxdamantus | since it just involves copying blocks across, which are usually 1GiB big. | 12:54 |
sicelo | mmm, my Thinkpad X40 is begging for retirement. in addition to all USB ports dead, since yesterday sound isn't working | 19:58 |
KotCzarny | poor thing.. | 19:59 |
KotCzarny | mine is still rolling strong | 19:59 |
KotCzarny | only plastic piece in one usb port broken | 19:59 |
KotCzarny | otherwise 100% functional, useful and kickin ass! | 19:59 |
KotCzarny | i got it 480gb ssd drive recently | 19:59 |
KotCzarny | so now it has 32gb ssd for os and 480gb for data | 20:00 |
sicelo | the problem with sound seems to be hardware, i guess, since linux picks up sound card but pulseaudio and alsa both can't produce any sound at all | 20:00 |
KotCzarny | try dos/windoze? | 20:00 |
sicelo | wowo @ 480GB SSD. mine has paltry 60GB | 20:01 |
KotCzarny | the 480gb one was on the sale, 80usd with free shipping | 20:01 |
KotCzarny | sh*tty chinese noname 3d nand tlc, but hey, it can hold a COPY of my data just fine | 20:01 |
sicelo | dunno if i'm up to the task of finding windows for the thing, as well as figure out how to boot it | 20:02 |
KotCzarny | https://www.geekbuying.com/item/Tigo-S300-480GB-Solid-State-Drive-2-5-Inch-SSD-Hard-Disk-398106.html | 20:02 |
KotCzarny | this one | 20:02 |
sicelo | mmm, unless i tftp boot it. | 20:02 |
sicelo | mine (X40) was given to me back in July 2016. the usb ports were already dead (evidently common x40 issue), and lid switch non-working | 20:03 |
sicelo | but it's been a nice little machine thus far :) | 20:04 |
KotCzarny | it still is | 20:04 |
sicelo | oh and bt, IR all non-working when i got it | 20:04 |
KotCzarny | same as n900, oldie, but goldie | 20:04 |
KotCzarny | maybe someone b0rked the firmware? | 20:04 |
KotCzarny | hard to imagine ir breaking | 20:05 |
KotCzarny | bt is a tiny addon on a modem card | 20:05 |
sicelo | i bought chinese pc card for usb, but it has never worked too well. works well only 2% of the time haha | 20:05 |
sicelo | looks like the laptop was once opened ... i have often wondered if there aren't loose connections inside or something (even for lid switch). just never got around to opening it to check. i probably might now | 20:07 |
sicelo | a couple of months ago i started having odd issue with MMC Card Reader - only reads card if i keep it pushed in :( | 20:08 |
sicelo | and batteries are finished too (lasting max 30 minutes now). | 20:09 |
sicelo | i can get battery ... but with the other issues, not sure if it's worth it | 20:09 |
sicelo | maybe if i open it, i can get bt & IR working :) | 20:10 |
KotCzarny | 30 minutes is a lot | 20:10 |
KotCzarny | my t500 has 30seconds | 20:10 |
KotCzarny | ;) | 20:10 |
sicelo | haha | 20:10 |
KotCzarny | but still enough for power surges so i dont change it | 20:11 |
KotCzarny | and if i'm quick enough i can change rooms | 20:11 |
sicelo | oh and hibernation doesn't work anymore ... either a kernel or driver issue .. couldn't really figure it out (kernel 3.16 hibernates, but not 4.x) | 20:12 |
KotCzarny | also, docking station is very nice for x40, makes all the cables on the back instead of sides | 20:12 |
KotCzarny | and i can put 2nd hdd in (even sata with proper ultrabay adapter) | 20:12 |
sicelo | i've been thinking about getting either an X230 or T430 to replace the X40 | 20:13 |
KotCzarny | thinkpads after lenove took over arent the same quality (it got worse with every series) | 20:14 |
KotCzarny | still, nice to have core i5 cpu | 20:14 |
sicelo | their price rnge here can just about match the size of my wallet. although recently i see the X250/T450 is good too (besided chiclet keyboard). it's slightly above what i am able to pay, but might be worth it in the long run | 20:15 |
* sicelo wishes he didn't have this X40 as a first laptop haha | 20:16 | |
KotCzarny | ;) | 20:16 |
sicelo | ruined me, now i don't think i can use anything else | 20:16 |
KotCzarny | love the kb | 20:16 |
KotCzarny | love the build quality | 20:16 |
KotCzarny | love the features | 20:16 |
sicelo | the trackpoint!! absolutely love it, and kb, yes indeed | 20:16 |
KotCzarny | yup, clit love too | 20:16 |
KotCzarny | btw. want me to find and send you x40 mobo? | 20:17 |
sicelo | that would be awesome, if it can work out | 20:18 |
sicelo | :) | 20:18 |
sicelo | it's when it comes to computing that i wish i was living in Europe | 20:21 |
KotCzarny | usa is better for that | 20:21 |
KotCzarny | and now china | 20:21 |
sicelo | here even simple mobile data plan will set you back quite a bit of money | 20:22 |
Vajb | sicelo: try ultimate boot cd with that laptop. There might be soundcard tests too | 20:22 |
sicelo | no one complains that manufacturers don't provide security updates, because no one wants to download them anyway :p | 20:23 |
Vajb | at least it has wide amount of hw tests available | 20:23 |
sicelo | Vajb: thanks. will look at it. hope i can booth that over tftp. it's the only way i can boot anything on it, unless i wipe the hdd maybe | 20:24 |
sicelo | although i am not ver hopefully. sound has been working all along. i can't fthom why it would stop suddenly. | 20:25 |
Vajb | ah it doesn't have cd rom drive... | 20:25 |
KotCzarny | :> | 20:25 |
KotCzarny | and usb fried | 20:25 |
KotCzarny | but iso can be booted from grub | 20:25 |
sicelo | yeah. it's like a bootlooping N900 with broken USb :p | 20:25 |
Vajb | i wouldn't touch grub.... | 20:26 |
KotCzarny | grub1 is simple and nice | 20:26 |
Vajb | i fiddled with it few days and decided to start fresh | 20:26 |
sicelo | my debian install has grub 2 | 20:26 |
KotCzarny | again, try grub1 | 20:26 |
Vajb | i think it was grub 2 too | 20:27 |
KotCzarny | grub1 has a very simple config | 20:27 |
sicelo | i will do so | 20:27 |
KotCzarny | grub2 probably can boot iso too, but might require different config | 20:28 |
sicelo | a quick search does show possible configs. i'll download UB cd tomorrow and try | 20:30 |
bleb | is this a good place to talk about community-developed phones like chen's thing | 21:44 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!