*** lennart has joined #crux | 00:25 | |
*** maxus has quit IRC | 00:33 | |
*** dinx has quit IRC | 00:36 | |
*** dinx has joined #crux | 00:36 | |
*** sepen has joined #crux | 00:56 | |
sepen | m0rning | 00:57 |
---|---|---|
*** ahmrahtcheer has joined #crux | 01:03 | |
*** prologic has quit IRC | 01:03 | |
teK | hello | 01:04 |
Romster | g'day | 01:05 |
ahmrahtcheer | o/ | 01:05 |
pitillo | good morning | 01:17 |
*** joacim has joined #crux | 01:29 | |
*** ahmrahtcheer has quit IRC | 01:34 | |
*** deus_ex has joined #crux | 01:39 | |
*** namenlos has joined #crux | 01:41 | |
*** pedja has quit IRC | 01:41 | |
*** deus_ex has quit IRC | 01:59 | |
*** deus_ex has joined #crux | 02:01 | |
*** mike_k has joined #crux | 03:18 | |
teK | mike_k | 04:13 |
*** dinx has quit IRC | 04:29 | |
mike_k | teK: teK | 04:54 |
teK | zabbix 1.6 is out | 04:56 |
mike_k | teK: thanks, but I'd stick to the 1.4 branch for another 2 months. Until new and shiny bugs will be cleaned out. | 04:57 |
mike_k | or should I blindly bump the version? =) | 04:58 |
pitillo | hey, is someone using e17 snapshots version? I hope he can contact me. | 04:58 |
teK | mike_k: I will test it :) | 05:00 |
mike_k | teK: thanks, I'll give a try too. But I'm sure there will be many new bugs in a meanwhile. | 05:01 |
teK | yes? | 05:11 |
mike_k | rephraze, please | 05:13 |
teK | you expect 1.6 to have many bugs? | 05:15 |
*** maxus has joined #crux | 05:17 | |
*** onestep has joined #crux | 05:18 | |
mike_k | That was true just after every new stable branch release. Why should it be different now? The worst sort of bugs affect database structure (after an update procedure). That is what I am afraid of. | 05:18 |
*** dinx_ has joined #crux | 05:18 | |
*** onestep has left #crux | 05:20 | |
*** dinx__ has joined #crux | 05:24 | |
*** treach has joined #crux | 05:31 | |
*** dinx_ has quit IRC | 05:36 | |
treach | http://bugzilla.kernel.org/show_bug.cgi?id=11382 ... oops. :> | 05:38 |
*** dinx__ has quit IRC | 05:42 | |
teK | yeah.. | 05:46 |
*** dinx has joined #crux | 05:52 | |
*** maxus has quit IRC | 05:54 | |
*** clickonce has joined #crux | 06:51 | |
clickonce | Hello guys | 06:51 |
clickonce | Would you do -P fix or -P ignore when running lilo fails because the Windows partition is misaligned? (It does not end on cylinder boundary) | 06:54 |
Romster | you mean lowercase -p | 06:59 |
Romster | Always try -P ignore first, as -P fix will re-write the partition table, possibly destroying all partitions on the disk. | 06:59 |
Romster | man lilo | 06:59 |
*** thrice` has joined #crux | 07:00 | |
Romster | i'd be wanting to dd the 512 bytes to a usb stick just in case. | 07:00 |
clickonce | It is uppercase -P :) | 07:01 |
clickonce | Done. | 07:02 |
clickonce | Let's hope this shit will work. Stupid Microsoft and Lenovo putting their partitions where they shouldn't be. | 07:02 |
clickonce | I do not dare reboot... :) | 07:04 |
clickonce | ouch... | 07:08 |
clickonce | 3 errors when booting Linux and then kernel panic.. | 07:08 |
rehabdoll | atleast you booted :) | 07:10 |
*** prologic_ has quit IRC | 07:11 | |
Romster | it is indeed lowercase as by the man page | 07:11 |
Romster | -p Require interactive entry of all passwords specified as "" in the configuration file. | 07:12 |
Romster | -P {fix|ignore|<global-option>} | 07:12 |
Romster | oh bugger it i got it reversed. | 07:12 |
Romster | it's not easy to see the difference. | 07:12 |
Romster | stupid font. | 07:12 |
treach | that's what you get for using gothic fraktur fonts. ;) | 07:14 |
Romster | hehe | 07:16 |
treach | lol -> http://german.about.com/library/gallery/blfoto_fraktur06.htm | 07:20 |
Romster | geez | 07:24 |
Romster | i could only read the translation ;) | 07:25 |
treach | of course, but that's not really the point. :> | 07:25 |
Romster | true, well that typeface is worse to read than what i got now. | 07:26 |
treach | you write a letter, lambasting a font as "Schwabacher Judish", and what font do you use in the letterhead..? :D | 07:26 |
*** tribb has quit IRC | 07:26 | |
*** tribb has joined #crux | 07:28 | |
Romster | ah... | 07:28 |
Romster | good point. | 07:28 |
treach | errm, I mean "Schwabacher Jewish". Not sure what "Schwabacher" refers to, but I'd guess it's some german province or something. :> | 07:29 |
treach | apparently some village. | 07:29 |
Romster | maybe you'd know better in them matters. | 07:32 |
treach | still, hilarious. | 07:32 |
*** deus_ex is now known as pedja | 07:38 | |
*** schniggi1 has joined #crux | 07:49 | |
*** schniggie has quit IRC | 07:49 | |
clickonce | Windows boots fine. | 07:51 |
clickonce | I'll just have to recompile my goddamn kernel. | 07:51 |
treach | why? | 07:51 |
clickonce | Because it panics due to wrong modules being modules. | 07:51 |
treach | ah | 07:52 |
clickonce | :) | 07:52 |
treach | simple rule. Do not use modules. :> | 07:52 |
clickonce | True =) | 07:52 |
clickonce | I'll make everything non-modular now. | 07:52 |
clickonce | Btw, if I change the partition types of sda6, sda7 and sda8 to Linux and LinuxSWAP will that automatically reposition the partitions that I do not touch inside fdisk? (I wouldn't like to realign sda1 and sda2 since windows is there) | 07:54 |
treach | no. | 07:55 |
clickonce | So they'll stay where they are? | 07:55 |
treach | yes, unless you're using some terribly broken tool. | 07:56 |
clickonce | fdisk :) | 07:56 |
*** jdolan_ has joined #crux | 07:56 | |
treach | realignment shouldn't happen unless you start deleting and recreating partitions. | 07:56 |
clickonce | Okay, nice. | 07:56 |
clickonce | I have backed up all my precious iTunes music though | 07:58 |
cruxbot | [contrib.git]: ethtool: New port. | 08:14 |
sepen | just weird that nobody maintains ethtool on contrib yet :) | 08:15 |
clickonce | Compiling the kernel brings back so many memories... | 08:20 |
go|dfish | From the ricer days! | 08:22 |
clickonce | :) | 08:22 |
* treach takes opensuse 11.1 b1 for a testrun in vbox. | 08:22 | |
clickonce | Nah, more from when I was 12 and a nervous wreck installing Linux =) | 08:22 |
treach | hah | 08:22 |
thrice` | treach: given up on 11.0? | 08:24 |
treach | no. "in vbox" :) | 08:24 |
thrice` | so, you're running 11.1 b1 from 11.0? | 08:24 |
thrice` | I read something funny about 11.1 b1. something with it frying network cards for people with e1000e | 08:25 |
*** kzk has left #crux | 08:26 | |
treach | thrice`: it's the kernel, so you might be careful with any .27 rc's.. | 08:27 |
treach | afk for a little while | 08:27 |
thrice` | heh, isn't ubuntu shipping that kernel as their default next month? :) | 08:27 |
dinx | http://lwn.net/Articles/299787/rss << | 08:30 |
clickonce | Damn kernel still panics. | 08:35 |
dinx | been through a few myself. i seem to have a few more modules than i should. | 08:35 |
clickonce | Hmm, it can't open my root partition. | 08:37 |
clickonce | Strangely I've included all the drivers that should be necessary. | 08:37 |
clickonce | At least I think I have. | 08:37 |
clickonce | Hmm, has /dev been changed to /devices? | 08:38 |
clickonce | Jeez, I really need to use Linux more often. | 08:38 |
clickonce | I am like a newbie all over again. | 08:38 |
dinx | O_o | 08:38 |
* Romster looks awestruck | 08:41 | |
* clickonce looks confused | 08:41 | |
Romster | /dev has not changed. | 08:41 |
clickonce | When my kernel boots, everything I see is /devices/platform paths. :) | 08:42 |
Romster | and i have very little issues with modules, noting a recompile and depmod -a don't fix. | 08:42 |
clickonce | I was begining to think that is why it can't find /dev/sda8 | 08:42 |
Romster | well you've broke something. | 08:43 |
clickonce | I guess I have the wrong drivers. | 08:43 |
Romster | lspci -nk | 08:43 |
clickonce | It's a ThinkPad T61, do I need SCSI device support for the kernel to find my hdd? | 08:46 |
dinx | depends wheather its full IDE or not | 08:46 |
dinx | if* | 08:47 |
clickonce | okay | 08:49 |
jaeger | or if you use the new libata drivers, which need scsi support | 08:50 |
clickonce | I said yes to what I need under SATA (Prod) PATA (Experimental). I do not know if that use libata or not. | 08:55 |
jaeger | that is indeed libata | 08:55 |
* clickonce goes to **** himself | 08:56 | |
clickonce | Damn, more than one year without compiling a kernel and you're back at 200BC | 08:56 |
*** miossyx has joined #crux | 08:57 | |
miossyx | Hi guys! | 08:57 |
jaeger | heyo, miossyx | 08:58 |
miossyx | Some boot-from-raid specialists here? | 08:58 |
miossyx | I want to install CRUX to a RAID1 but I cannot get lilo to boot it. | 08:59 |
miossyx | Hi, jaeger! | 08:59 |
mike_k | real hw raid? | 08:59 |
miossyx | nope.. | 08:59 |
miossyx | sw raid. | 08:59 |
miossyx | two sata harddisks on Promise TX2 300 | 08:59 |
miossyx | md0 = sda1 + sdb1 | 08:59 |
miossyx | crux is on md0... | 09:00 |
jaeger | have a little experience there though haven't done it with crux specifically | 09:00 |
miossyx | Well... the raid works fine | 09:00 |
jaeger | not too hard, just need to set up an initrd with enough to get your raid going before the rest of the boot | 09:01 |
miossyx | I just cannot tell lilo to boot from /dev/md0 | 09:01 |
clickonce | Ramdisks are hell... | 09:01 |
miossyx | no initrd... | 09:01 |
mike_k | tell it to boot from sda | 09:01 |
mike_k | and make sure you have a working boot loader on sdb too (to boot if the first hdd fail) | 09:02 |
jaeger | yeah, lilo doesn't have to know about the raid, it can boot from one disk or the other and then the raid takes over after | 09:02 |
miossyx | I had a setup like this (without initrd) running in my former company... but don't remember how I configured lilo there. | 09:02 |
miossyx | Jup... lilo goes to both hdds | 09:02 |
miossyx | Hold on a second... I'll post the config. | 09:02 |
mike_k | you may also install a boot loader on a usb-stick | 09:02 |
jaeger | or a floppy or cd if that's the way you prefer | 09:03 |
mike_k | I have a running crux on mdraid somewhere... without a damn initrd though | 09:03 |
miossyx | boot=/dev/md0 | 09:03 |
miossyx | root=/dev/md0 | 09:03 |
miossyx | raid-extra-boot=auto | 09:03 |
miossyx | lilo runs fine... kernel 2.6.26.5 is used. | 09:04 |
miossyx | but it doesn't boot... | 09:04 |
miossyx | sda1 and sdb1 are marked as bootable. (fdisk flag) | 09:04 |
jaeger | without an initrd your root raid isn't configured yet at the point it tries to switch the root to /dev/md0, I would think | 09:05 |
treach | and lilo doesn't understand md's anyway iirc | 09:05 |
miossyx | nope. | 09:05 |
treach | not sure if grub does | 09:05 |
mike_k | grub: kernel (hd0,0)/boot/vmlinuz root=/dev/md0 | 09:06 |
clickonce | After doing make all it said "root device is (8,8)" does that mean it has somewhat selected the root= device during compilation? | 09:06 |
treach | "make help" | 09:06 |
treach | do not just blindly use arguments to make for the kernel | 09:07 |
clickonce | I haven't. | 09:07 |
clickonce | "make all" does do what I want to do. I just haven't ever seen it say that "root device is" after a compilation. | 09:07 |
treach | ok | 09:08 |
clickonce | make all does: make vmlinux modules bzImage | 09:08 |
miossyx | Hmm... I try it with lilo -m /dev/sda | 09:08 |
mike_k | miossyx: simply install grub (I don't know why it should fail with lilo) on each of hdd's like it is the first one (changing device.map accordingly http://forums.fedoraforum.org/showthread.php?t=153679) and boot it from either hdd. | 09:08 |
miossyx | Nope... | 09:09 |
* treach is back at trying to trick 11.1 into vbox | 09:09 | |
miossyx | I want lilo... | 09:09 |
miossyx | I already had a setup where lilo is working... | 09:09 |
miossyx | lilo -M /dev/sda to be precise. | 09:09 |
*** Phil-san has joined #crux | 09:10 | |
miossyx | Ok... | 09:11 |
miossyx | now it works... | 09:11 |
jaeger | treach: 11.1 of what? | 09:12 |
treach | beta 1, opensuse | 09:12 |
jaeger | ah | 09:12 |
miossyx | Anybody interested in "how to install CRUX on RAID1"? | 09:12 |
miossyx | Maybe even for the website? | 09:13 |
treach | already found some interesting bugs. apparently the installer won't let you create a swap partition, but when you try to proceed without one, it warns you and says something like "do you really want to proceed like this?" :p | 09:13 |
jaeger | miossyx: can't hurt, in my opinion. that's what the public wiki is for, no? | 09:14 |
miossyx | Hmm... I'll have to check my account there... | 09:14 |
miossyx | I've been away from crux for a while. | 09:15 |
jaeger | I know the feeling | 09:17 |
jaeger | I gave up most of my ports because I lost the time to dedicate to crux, only recently coming back | 09:17 |
treach | crux, addictive substance for computer geeks. :> | 09:17 |
jaeger | heh | 09:17 |
miossyx | Well, I've changed my job... unfortunately to a Windows-only company. | 09:17 |
miossyx | I am the first here with a linux server. | 09:18 |
jaeger | gonna try to get an install of it going on my eee soon | 09:18 |
miossyx | Well... and it's CRUX of course ;-) | 09:18 |
treach | jaeger: that sounds.. interesting. :p | 09:18 |
miossyx | I got an aspire one last week... | 09:18 |
jaeger | treach: at the moment I have ubuntu-eee on it, pretty fun little box | 09:18 |
treach | yeah, I'm envious. :) | 09:19 |
jaeger | have /usr and /var/cache compressed with squashfs and overlayed with a CoW type unionfs on both | 09:19 |
treach | I wish I had the budget to get one of those little buggers. :> | 09:19 |
dinx | jaeger: you still looking after gnome ports ? | 09:19 |
jaeger | dinx: yes, though they're in need of some updating, which I get too slowly | 09:20 |
treach | compiling firefox on it seems like a bit of a challenge though. ;) | 09:20 |
jaeger | how so? | 09:20 |
miossyx | Basically, I can recommend the Acer Aspire One. The only thing which is really bad is the audio output... | 09:21 |
treach | ssd disks, cpu adequate, but not really designed for tasks like that, I guess. | 09:21 |
jaeger | treach: ah | 09:21 |
miossyx | 120GB HDD, 1GB RAM, ... | 09:21 |
jaeger | I'd probably build it in a tmpfs if possible | 09:21 |
miossyx | quite nice. | 09:21 |
jaeger | 20GB SSD and 2G RAM on mine | 09:21 |
miossyx | 3h with a full battery | 09:21 |
* dinx hates dinx. since they stuffed a Broadcom card in his aspire 5100 | 09:22 | |
dinx | eh | 09:22 |
dinx | hates Acer | 09:22 |
jaeger | build in a tmpfs or put a 1.8" HD in it. there's room in it, though not a lot | 09:22 |
miossyx | Another bug: If the backlight brightness is turned low, it starts to flicker when there is access to the harddisk... :-( | 09:22 |
jaeger | doh, that's weird | 09:22 |
miossyx | With a new BIOS they just limited the brightness to the brighter three steps. | 09:23 |
jaeger | dinx: I had tons of trouble getting my aspire 5150's atheros card to work but with a very specific madwifi snapshot it's been happy | 09:23 |
miossyx | But the wlan on this little thing is fascinating... | 09:23 |
dinx | jaeger: cool. bought a PCMCIA Atheros based and had no troubles since | 09:24 |
dinx | untill the day i dropped my laptop and cracked the card | 09:24 |
dinx | so i am back to the broadcom one | 09:24 |
miossyx | I can even see ssids from other houses 200m away in downtown munich. | 09:24 |
dinx | using ndiswrapper | 09:24 |
dinx | im sitting in the same room as router and signal strength is 70% O_o | 09:24 |
jaeger | dinx: ah, that sucks :( | 09:25 |
jaeger | miossyx: nice | 09:25 |
treach | "Free Internet" :p | 09:25 |
dinx | miossyx: sweet. | 09:25 |
dinx | treach: lol | 09:25 |
treach | heh, my phone picks up my neighbours wlans, but I haven't had time playing with it. Hopefully it doesn'a announce itself on their nets without telling me. :> | 09:27 |
clickonce | Wow, my screen goes crazy when I start Xorg. | 09:28 |
clickonce | White, blue and green lines appear here and there. | 09:29 |
miossyx | Well... how can somebody register at the CRUX wiki? | 09:29 |
jaeger | hrmm, odd... my eee drained its battery while suspended | 09:39 |
treach | it does that apparently | 09:40 |
jaeger | good to know | 09:40 |
treach | it seems to be a known problem, apparently the battery leaks, about 14mA even when the system is *off*.. | 09:42 |
*** Phil-san has quit IRC | 09:43 | |
jaeger | I guess I'll plug it in more often when I'm not using it mobile, hehe | 09:43 |
treach | :) | 09:43 |
treach | heh, seems like the "Wind" have the simliar issues. :p | 09:45 |
Romster | lilo can boot from RAID-1 only. | 09:46 |
miossyx | romster: Yes it can. :-) | 09:46 |
jaeger | even with that it's still a pretty cool box, I don't regret buying it :) | 09:48 |
treach | yeah. "traditional" laptops are just too heavy lugging around. | 09:49 |
treach | either that, or you have to rob a bank first, before purchase. :> | 09:50 |
jaeger | indeed | 09:50 |
jaeger | heh | 09:50 |
Romster | and guard it with your life. | 09:51 |
treach | yep | 09:51 |
treach | I had a real scare when I forgot my r40 at the uni a few years ago.. A minor miracle I didn't run someone over on my way back. :> | 09:54 |
Romster | geez | 09:57 |
clickonce | WIe | 10:06 |
clickonce | I got it running. | 10:06 |
*** fetid has joined #crux | 10:09 | |
miossyx | http://crux.nu/Wiki/LILOBootFromRaid1 | 10:13 |
miossyx | How can I add it to a Category to have it shown on http://crux.nu/Wiki/HomePage ? | 10:15 |
*** RedShift has joined #crux | 10:23 | |
cruxbot | [xfce.git]: libwnck: Removed. Moved to opt collection. | 10:23 |
*** jue has joined #crux | 10:26 | |
*** ChanServ sets mode: +o jue | 10:26 | |
cruxbot | [opt.git]: libwnck: New port (moved from xfce and gnome collections) | 10:26 |
sepen | jaeger, now Im maintaining libwnck in opt, and I removed it from xfce, so if you want ... feel free to remove it from gnome collection too | 10:26 |
*** dinx has quit IRC | 10:26 | |
*** ximonda has joined #crux | 10:26 | |
ximonda | Ok... back with my last nickname... | 10:26 |
miossyx | miossyx -> ximonda. | 10:27 |
*** miossyx has left #crux | 10:27 | |
ximonda | better. | 10:27 |
ximonda | Hmm... can anybody tell me how I can add my Wiki page to a Categroy so it's shown on http://crux.nu/Wiki/HomePage ? | 10:28 |
jue | just add something like -> Categories: [[!Install]] | 10:29 |
ximonda | to my page? | 10:29 |
jue | yep | 10:29 |
ximonda | Ok, thank you. | 10:30 |
jue | np | 10:30 |
sepen | bbl, going to home | 10:30 |
*** sepen has quit IRC | 10:30 | |
ximonda | ok... | 10:32 |
ximonda | Please have a look and check for typos... | 10:33 |
treach | "harddisks", I presume. | 10:34 |
treach | ximonda: besides, there is no point in creating raid arrays for swap | 10:36 |
ximonda | Hmm... I've read something different. | 10:37 |
treach | well. that was wrong. | 10:38 |
ximonda | If a disk fails, the system doesn't crash if swap is on raid. | 10:38 |
ximonda | and swap on raid1 is tested well. | 10:38 |
cruxbot | [opt.git]: sqlite3: update to 3.6.3 | 10:39 |
cruxbot | [opt.git]: lame: update to 3.98.2 | 10:39 |
cruxbot | [opt.git]: xsane: update to 0.996 | 10:39 |
treach | ....ok. Personally I'd question if swap was valuable enough to warrand a raid1 partition, but what ever floats your boat. | 10:39 |
treach | s/warrand/warrant/ | 10:40 |
ximonda | Hmm... For this server I really prefer uptime and reliability over anything else... | 10:41 |
treach | besides, there's still a 50% risk it'll go down anyway. :> | 10:41 |
ximonda | And a failing swap parition will most likely kill your system... | 10:41 |
ximonda | trach: please explain... | 10:41 |
treach | a redundat swap won't save you if you're swapping on the faulty drive. | 10:41 |
ximonda | swap is on /dev/md2 | 10:42 |
treach | *redundant*, dammit | 10:42 |
treach | yes,yes | 10:42 |
ximonda | of one disk fails, the system shouldn't go down. | 10:42 |
treach | but it's made up of two parts | 10:42 |
ximonda | raid1! | 10:42 |
treach | indeed. | 10:42 |
treach | but how does raid1 work, really? | 10:42 |
ximonda | Sorry, I don't get your point here. | 10:42 |
treach | usually you just use one disk and mirror it to the other. | 10:43 |
*** dinx has joined #crux | 10:43 | |
*** mrks has joined #crux | 10:43 | |
treach | note that I'm not trying to shoot down the idea, just pointing out that there's more to it than just creating a mirror and then calling it a day. | 10:44 |
ximonda | So, you think that if the system is in heavy swap on /dev/md2, it will break down if one harddisk fails? | 10:46 |
treach | no, I say there's a 50% risk it does | 10:46 |
ximonda | What's the point of raid1 then? | 10:47 |
ximonda | Swap = data... | 10:47 |
ximonda | so if your statement would be correct, there would be no point in putting data on a raid1 | 10:47 |
treach | you're confusing different things. | 10:48 |
ximonda | it it would fail with 50% on a single hdd failure. | 10:48 |
ximonda | if... | 10:48 |
treach | your data != your swap. | 10:48 |
ximonda | nope. | 10:48 |
treach | but anyway, relying on raid 1 instead of backup is foolish anyway. | 10:48 |
ximonda | Well... there are backups... anyway. | 10:49 |
ximonda | I just want to save maintenance time if a harddisk fails. | 10:49 |
treach | well. it's not like you can leave it that way, right? :) | 10:49 |
*** mike_k has quit IRC | 10:50 | |
treach | considering the hypothetical corrupt swap partition is on the same failing drive as some of your data. :) | 10:50 |
ximonda | Quoted from http://www.linuxwiki.de/LinuxSoftwareRAID | 10:51 |
ximonda | wer Abstürze vermeiden will, swappt auf RAID1, dann steht der Swapspace auch noch bei Plattenausfall zur Verfügung | 10:51 |
ximonda | * Swappen auf RAID1 ist praxisgetestet und funktioniert spätestens ab Kernel 2.4.17 vorzüglich und auch bei hoher Last sowie bei parallel laufendem RAID-Resync! | 10:51 |
ximonda | Only in german... | 10:52 |
ximonda | So... that's what I want... prevent downtime. | 10:52 |
treach | tut mir leid, aber ich glaube daß mein deutsch nicht sharf genüg for das ist. :> | 10:52 |
treach | aber, anyway. 50% risk of failure is still better than 100%, right? :) | 10:53 |
ximonda | Translation: If you want to prevent crashes, swap on RAID1... swapspace is still available when a harddisk goes down. | 10:53 |
treach | yes, but it's an oversimplification. | 10:53 |
ximonda | swap on RAID1 is tested and working since 2.4.17 perfectly and even during heavy load and during a running resync pf the raid. | 10:54 |
treach | yesyes, | 10:54 |
treach | that's not what I'm saying. | 10:54 |
ximonda | Well... I just don't see your 50% are coming from... | 10:54 |
treach | I'm not saying it doesn't work. | 10:54 |
ximonda | Maybe I had too much beer yesterday evening at the Octoberfest. ;-) | 10:55 |
treach | ximonda: I suggest you take a good look at how exactly raid 1 works. | 10:55 |
treach | which disks are used, and how. | 10:55 |
ximonda | Ok, treach... I will read the code of md. | 10:55 |
treach | it's NOT like stuff gets written simultaneously on both disks. | 10:56 |
treach | well, that might be taking it a bit too far. :p | 10:56 |
ximonda | And anyway... I did several tests on some other server also with swap on raid... no problems at all when I unplugged one of the harddisks.. | 10:57 |
treach | I do not get where you got the idea that I claimed so. | 10:57 |
ximonda | Sorry, I still don't get your point... | 10:58 |
treach | it's a probability problem. anecdotical evidence is useless. | 10:58 |
treach | ximonda: my point have to do with how raid 1 really works, and how that relates to swap, and corrupt data. | 10:59 |
treach | you do know that it is possible for a raid 1 to mirror corrupt data, if you're unlucky? | 10:59 |
ximonda | I am quite sure how raid1 works... I was hacking quite some time in linux kernel drivers... | 10:59 |
treach | great. | 10:59 |
ximonda | Data is written to both disks from memory... | 11:00 |
ximonda | and not read back and mirrored to the other disk... | 11:00 |
ximonda | that's old style raid1. | 11:00 |
ximonda | If it would be as you wrote above, you would be right... | 11:01 |
treach | I know it's written from memory. but how is it read, and which disk is the "master"? | 11:01 |
ximonda | Well... | 11:01 |
ximonda | Sorry, I don't want to discuss this further... | 11:01 |
treach | fine with me, I feel we are not really getting eachother. | 11:02 |
ximonda | Well... doesn't matter. | 11:02 |
treach | "Wir sprechen einander vorbei", I think you'd say. :) | 11:02 |
ximonda | A kind of... | 11:02 |
ximonda | I think understood your point now... but I am quite sure that linux' raid1 works differently. | 11:03 |
ximonda | Well... reading the code will tell... | 11:03 |
treach | ok. | 11:04 |
ximonda | or asking Andrew | 11:04 |
ximonda | or somebody from the raid ml | 11:04 |
*** sepen has joined #crux | 11:06 | |
treach | ximonda: according to http://linas.org/linux/Software-RAID/Software-RAID-8.html | 11:09 |
treach | "MD implements read balancing. That is, the RAID-1 code will alternate between each of the (two or more) disks in the mirror, making alternate reads to each." | 11:09 |
treach | ie, one good disk won't save you, since half the swap would be corrupt. | 11:09 |
ximonda | I would expect md to realize a read error from the disk and rereads the data from the other disk. | 11:11 |
Romster | i use raid on 2 disks with the same priority and it shares out between them ximonda | 11:11 |
Romster | http://pastebin.ca/1209425 | 11:12 |
Romster | err s/raid/swap/ | 11:12 |
treach | ximonda: true in the best case. (drive is dead.) | 11:12 |
ximonda | When I tried to simulate failing harddisks (hot-unplugging them) at some time around 2.6.18 the md took about 20 seconds to realize that a disk was failing (throwing tons of errors). | 11:12 |
ximonda | But then it recovered fine without further complains. | 11:13 |
treach | yeah, again, I'm not saying it doesn't work. | 11:13 |
ximonda | In the server I just replacing the disks I had a harddisk with read-errors... | 11:13 |
Romster | i've never tired to hot swap though | 11:13 |
ximonda | When the data came back, the data was correct... | 11:14 |
treach | Romster: with ATA? :D | 11:14 |
ximonda | when the harddisk couldn't read the data, it just threw an error... | 11:14 |
Romster | yeah sata2 | 11:14 |
Romster | the hardware probably isn't hot swapable. | 11:15 |
ximonda | I hot-plugged pata... it's rough but no problems there. | 11:15 |
Romster | although i have pulled out cdroms on IDE in the past :D | 11:15 |
ximonda | sata is electrically hot-pluggable. | 11:15 |
ximonda | sas can also do it logically. | 11:15 |
Romster | ximonda, but if the driver handles it is the issue. | 11:15 |
aon | i once accidentally hotunplugged pata and was wondering why the hell new software wouldnt't start | 11:15 |
aon | solaris9 didn't crash from detaching the hdd | 11:15 |
Romster | sweet. | 11:16 |
Romster | jsut ran off ram eh | 11:16 |
treach | heh, sata should handle hotplugging.. question is, is it correctly implemented by everyone involved? :> | 11:17 |
treach | I've never dared trying. :p | 11:17 |
Romster | i haven't had the need to yet. | 11:17 |
ximonda | Linux doesn't crash if you unplug a harddisk and re-plug it... | 11:18 |
treach | ximonda: that's not what I mean. | 11:18 |
Romster | also been meaning to test powering down some of my cpu cores to sleep. | 11:18 |
ximonda | The drivers reinitialize the harddisks when it was detached as part of their error-recovery. | 11:18 |
treach | ximonda: I'm just afraid I'll be standing in a cloud of blue magic smoke if I try because some little korean rice peeler screwed up someplace. :> | 11:19 |
*** maxus has joined #crux | 11:19 | |
Romster | nah at worse you'll end up with a dead controller :D least would be a freeze. | 11:20 |
ximonda | Well... I had to make sure the system keeps on running at that time... otherwise I would have never got on holiday. | 11:20 |
ximonda | Ok, guys, I have to leave now.. going home. | 11:21 |
Romster | later. | 11:21 |
treach | take care | 11:21 |
ximonda | Thank you for the discussion, treach. :-) | 11:21 |
treach | heh, you're welcome. :p | 11:21 |
ximonda | And thanks for the help anyway :-) | 11:21 |
ximonda | cu | 11:21 |
*** ximonda has left #crux | 11:22 | |
*** fetid has quit IRC | 11:23 | |
*** roliveira has joined #crux | 11:54 | |
*** namenlos has quit IRC | 12:00 | |
*** dinx has quit IRC | 13:01 | |
clickonce | Hmm, "Network Controller: Intel Corporation Unknown device 4230" must be my wireless intel card, mustn't it? My 1Gbit Intel card is already working. | 13:07 |
jaeger | possibly | 13:09 |
jaeger | get any more info if you update the pci IDs? | 13:09 |
*** dinx has joined #crux | 13:21 | |
sepen | jaeger, did you see my comments about libwnck? | 13:47 |
jaeger | yes | 14:07 |
jaeger | I wasn't here at the time but I saw it later | 14:07 |
sepen | ok | 14:11 |
clickonce | Hmm, would it be possible to get dualhead console? Like switch between monitors using keycombination X and switch between virtual consoles on that monitor using the regular Meta-FX | 14:14 |
sepen | clb, did you know 'screen'? | 14:15 |
sepen | opps clickonce | 14:15 |
sepen | dualhead console? :) my bad | 14:16 |
jaeger | never seen a thing like that, myself, but it would be neat | 14:22 |
*** maxus has quit IRC | 14:22 | |
*** mrks has quit IRC | 14:26 | |
clickonce | Hmm | 14:38 |
*** maxus has joined #crux | 14:39 | |
clickonce | I did include the generic IEEE802.11 stack and all Intel PRO/Wireless drivers. It still doesn't find my network card. | 14:40 |
clickonce | Which is a 4965AGN | 14:41 |
clickonce | Do I need the stuff from intellinuxwireless anyway? | 14:41 |
clickonce | I cannot find anything specifically 4965AGN in make menuconfig though. | 14:42 |
clickonce | Ah, had to turn on showing developement/incomplete stuff. | 14:46 |
Ovim-Obscurum | hi! | 14:52 |
Ovim-Obscurum | pcmcia card? | 14:52 |
Ovim-Obscurum | or usb/pci? | 14:52 |
Ovim-Obscurum | brb | 14:55 |
*** onestep has joined #crux | 14:58 | |
clickonce | Ovim-Obscurum: PCI | 14:59 |
thrice` | you need to install the firmware | 15:06 |
*** roliveira has quit IRC | 15:19 | |
*** sepen has quit IRC | 15:26 | |
*** sepen has joined #crux | 15:28 | |
Ovim-Obscurum | clickonce: is it listed in lspci? | 15:33 |
clickonce | It's up and running now. | 15:36 |
Ovim-Obscurum | Ah, haven't seen, was afk ;) | 15:36 |
Ovim-Obscurum | Nice :) | 15:36 |
clickonce | You haven't missed anything. :) | 15:36 |
Ovim-Obscurum | :D | 15:36 |
clickonce | Now I just have to get wpa_supplicant configured and I'm all set. | 15:37 |
Ovim-Obscurum | Congratulations :D | 15:37 |
clickonce | Thank you! =) | 15:37 |
Ovim-Obscurum | no problem :) | 15:37 |
*** DarkNekros has joined #crux | 15:41 | |
*** onestep has quit IRC | 16:01 | |
*** roliveira has joined #crux | 16:03 | |
*** jdolan_ has quit IRC | 16:16 | |
*** jue has quit IRC | 16:45 | |
clickonce | Hmm, my ThinkPad | 16:53 |
clickonce | Hmm, my ThinkPad's fan makes this high-pitched noise when being disconnected from the AC adapter. | 16:53 |
clickonce | It does not matter wether I run Windows or any other operating system. | 16:53 |
treach | looks like you're in the wrong channel then. ;) | 16:55 |
clickonce | Hmm, you're right. | 16:56 |
* treach puts on his most patronizing face; "I usually am." | 16:57 | |
treach | :p | 16:57 |
clickonce | bah :P | 16:58 |
clickonce | I put it in the window to the left of #crux, much better. (##hardware) | 16:58 |
treach | hah, ok. :) | 16:59 |
treach | I used to stack the channels within irssi, but I had to quit since I frequently spoke in the wrong channel. :> | 17:00 |
clickonce | I have my windows occupy the entire main window. | 17:01 |
*** RedShift has quit IRC | 17:02 | |
*** ahmrahtcheer has joined #crux | 17:09 | |
*** treach has quit IRC | 17:15 | |
*** prologic has joined #crux | 17:18 | |
*** ahmrahtcheer has quit IRC | 17:20 | |
*** joacim has quit IRC | 17:29 | |
*** lennart has quit IRC | 17:35 | |
*** prologic has joined #crux | 18:08 | |
*** roliveira has quit IRC | 18:11 | |
*** racer has joined #crux | 18:58 | |
*** dinx has quit IRC | 19:05 | |
*** nipuL has quit IRC | 19:30 | |
*** nipuL has joined #crux | 19:31 | |
*** sepen has quit IRC | 19:36 | |
*** racer has quit IRC | 19:48 | |
nipuL | omg, one of my clients actually paid me | 20:29 |
*** racer has joined #crux | 20:31 | |
prologic | for what ? | 20:42 |
nipuL | services rendered | 20:45 |
*** Dudde has quit IRC | 20:51 | |
*** mavrick61 has quit IRC | 20:51 | |
*** mavrick61 has joined #crux | 20:52 | |
*** Dudde has joined #crux | 20:53 | |
*** prologic has quit IRC | 21:51 | |
*** racer has quit IRC | 21:51 | |
*** fetid has joined #crux | 21:59 | |
*** prologic has joined #crux | 22:33 | |
*** morlenxus has quit IRC | 23:05 | |
*** morlenxus has joined #crux | 23:05 |
Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!