Web   ·   Wiki   ·   Activities   ·   Blog   ·   Lists   ·   Chat   ·   Meeting   ·   Bugs   ·   Git   ·   Translate   ·   Archive   ·   People   ·   Donate

#olpc-paraguay, 2010-10-20

 « Previous day | Index | Today | Next day »     Channels | Search | Join

All times shown according to UTC.

Time Nick Message
00:44 tch has quit IRC
00:45 tch <tch!~tch@> has joined #olpc-paraguay
00:57 tch has quit IRC
00:59 tch <tch!~tch@> has joined #olpc-paraguay
01:05 tch has quit IRC
01:05 tch <tch!~tch@> has joined #olpc-paraguay
01:12 tch has quit IRC
01:27 tch <tch!~tch@> has joined #olpc-paraguay
01:32 tch has quit IRC
01:38 manusheel has quit IRC
01:41 manusheel <manusheel!~manusheel@gateway/shell/sugarlabs.org/x-zbexbkwigghqmjmz> has joined #olpc-paraguay
01:43 tch <tch!~tch@> has joined #olpc-paraguay
01:47 tch has quit IRC
01:47 tch <tch!~tch@> has joined #olpc-paraguay
01:51 tch has quit IRC
01:51 tch <tch!~tch@> has joined #olpc-paraguay
01:55 tch has quit IRC
03:19 tch <tch!~tch@> has joined #olpc-paraguay
07:17 manusheel is now known as manusheel_afk
08:47 manusheel_afk is now known as manusheel
10:56 Richar_ <Richar_!~Paraguay@mail.paraguayeduca.org> has joined #olpc-paraguay
11:03 Richar_ bernie: hello
11:03 bernie: me podrias pasar el mail walter bender
11:04 bernie: quede en pasarle links de trabajos a su correo
11:07 bernie: ya lo consegui igualmente gracias
11:07 Richar_ has quit IRC
11:37 bernie damn
11:37 he left before I could answer
11:41 rgs__ bernie: he got the address anyways
11:42 bernie rgs__: be careful not to loopback maount while the vm is running or you'll corrupt the fs
11:42 rgs__ bernie: VM is dead
11:42 bernie: any thoughts on the apparmor error msg?
11:42 would you care aobut it?
11:42 bernie rgs__: nope
11:42 rgs__: btw, apparmor is even more pointless than selinux. I don't know why I did not disable it
11:42 rgs__: I believe in security, but not in that sort of crap
11:43 rgs__: lol, you caused a munin alert! :-)
11:44 rgs__ bernie: because of the mount?!?
11:45 bernie rgs__: yes, munin immediately reported that it does not know its size. stupid munin
11:45 rgs__: *CONTEXT_SWITCH* I've replaced the naughty gnome-terminal with terminator. it's written in python, but it works very well.
11:46 rgs__ bernie: hehehe. really? Does it handle multi-tab ok?
11:46 bernie rgs__: it also does vertical and horizontal splits
11:47 rgs__: and has great default shortcuts to control the views
11:47 rgs__ bernie: can dmesg's buffer ring be full?
11:47 bernie: running dmesg on housetree gives me half-printed messages. weired.
11:47 *weird
11:47 bernie: its a ring.. I'd suppose it would just overwrite itself ...
11:48 bernie: nevermind , housetree is just being slugish
11:51 bernie reubencaron_: yes, dmesg has a limited ring buffer. it's configurable at compile-time. often 64kb
11:51 rgs__: ^^^ oops
11:51 rgs__ bernie: partition looks ok..
11:51 bernie: maybe we could run on fsck, can we feed an off-set to fsck ?
11:53 bernie rgs__: from the boot messages, it looks like it's hanging before loading initrd
11:54 rgs__: why don't you do this: extract kernel and initrd to a /srv/vm/kernel/...
11:54 rgs__: then change the .xml to make it load the kernel and initrd externally. like many other vms do.
11:54 rgs__ bernie: k, will look how to explicitly set which kernel to boot
11:54 bernie rgs__: I do not like the partitioned volumes with grub in them... I prefer to specify the kernel/initrd externally and use regular partitions for booting
11:55 rgs__: the partitioned lv is the result of how image-create clones the template vm
12:17 rgs__ bernie: you rock
12:17 bernie: using an external kernel did it
12:18 bernie wow
12:18 rgs__ bernie: question is, why did the internal kernel stopped working?
12:19 bernie: would you be so nice to re-assign me an IP for serendipity (I'll tell you which one it is)? Can't push myself because I didn't fix the problem with bind-utils that is not allowing me to sign zones
12:20 bernie:
12:20 bernie: --> serendipity.sl.o
12:20 bernie: I'll run a full-upgrade and test our luck again with booting from a partition
12:20 bernie rgs__: i have no idea why it stopped working
12:20 rgs__: I just don't like that sort of setup...
12:21 rgs__ bernie: me neither. Which sort?
12:21 bernie rgs__: now maybe you could siphon the files from the odd partition to a real lvm partition
12:21 rgs__ bernie: (please push the DNS change so I can secure the IP address)
12:21 bernie rgs__: the setup with an MBR partitioned disk inside an lvm partition. it's redundant and awkward.
12:21 rgs__: also, it makes it impossible to grow the partition on the fly
12:21 jasg <jasg!~jorge@mail.paraguayeduca.org> has joined #olpc-paraguay
12:21 rgs__ bernie: errr.. later, this detour took lots of time. I need to get back to my main thread
12:22 bernie rgs__: why can't you push the dns change yourself?
12:22 ah... I got it. the dnssec problem
12:22 rgs__ bernie: can't sign :(
12:22 bernie: will fix that later too.. (damn maverick)
12:22 bernie rgs__: if you're like me, the "later" things never get done
12:23 car|0s <car|0s!~carlos@> has joined #olpc-paraguay
12:23 rgs__ bernie: it will. When I get the need to sign something and you are not around
12:23 bernie rgs__: I have 2 years old devtodo entries such as "cancel mastercard in italy"
12:23 rgs__ bernie: hahahaha
12:23 bernie: I have similar age TODOs too
12:23 car|0s rgs__, bernie morning!
12:24 rgs__ car|0s: Carloncho
12:24 car|0s rgs__, que feo suena eso
12:24 rgs__ car|0s: no se porque pense en britania cuando vi tu saludo.. que rara asociacion!
12:24 car|0s rgs__, @ CLEI
12:24 rgs__, jajaja
12:24 rgs__ car|0s: que bueno!
12:25 bernie: pushed? (sorry - i really need it!)
12:25 bernie rgs__: not yet
12:25 rgs__: for some reason i have conflicts
12:26 rgs__ bernie: ouch, hope it wasn't me
12:26 bernie: the last change I have is from Oct the 13th.. and its mine
12:26 r
12:26 commit ec041ba1d9c55c8377311bf5138c5d035944a265
12:26 Author: Raúl Gutiérrez S <rgs@rieder.net.py>
12:26 Date:   Wed Oct 13 17:10:36 2010 +0200
12:26 Releasing the IP address of an unused VM.
12:28 bernie rgs__: yes, that was the change. the conflict was because somehow I had unpushed changes. not your fault!
12:29 rgs__ bernie: best software available for blogging? Wordpress?
12:29 bernie rgs__: ok change pushed
12:29 rgs__: we have no reverse
12:29 rgs__: yes, wordpress is thebest
12:29 rgs__ bernie: not a problem at the time..
12:29 bernie: hmmm.. checking up if there is anything cool done in RoR
12:29 bernie rgs__: btw, we have a wordpress-mu instance on sunjammer. icarito runs it
12:30 rgs__ bernie: this will serve me us a testbed, so I rathe be on my own
12:30 *rather
12:30 bernie: hey, how did you resolve the conflict?
12:30 bernie: believe or not, me never solved conflicts in Git 8-S
12:30 bernie: i am very amateur user
12:31 bernie rgs__: ah, would you like to do something for paraguay? I have a list of all the blogs of caacupe's teachers. we should setup py.planet.sugarlabs.org akin to the current planet.sugarlabs.org
12:31 rgs__ bernie: later :)
12:31 bernie: really, I have to play with RoR 3
12:31 bernie rgs__: I go to the conflicting file, see what's the matter, merge manually, then do "git add file" and then "git commit"
12:31 rgs__ bernie:  bought 3 uber cool ebooks yesterday. And today a friend told me he download them from a pirate site (ouch, it was 50 bucks)
12:31 bernie rgs__: instead of commit, it could be "git rebase --continue" or stuff like that
12:32 rgs__: which books? amazon links please
12:32 rgs__ bernie: but git nows you solved a conflict. What if you decided to ignore? Ah, then you would just revert your stuff right?
12:32 bernie: (they weren't from amazon - pragprog.com (Pragmatic Programmers)
12:32 bernie: links:
12:33 http://pragprog.com/titles/rai[…]opment-with-rails
12:34 http://pragprog.com/titles/cfc[…]ionate-programmer (really good, almost finished)
12:34 http://pragprog.com/titles/ruby/programming-ruby
12:34 bernie rgs__: I could do "git checkout ..." of the version I want to keep and then add it to say that  this is how I wanted to resolve the conflict
12:35 rgs__: but there's rarely a need to checkout a pristine copy, because the conflicted file contains copies of both versions (ours and theirs)
12:37 rgs__ bernie: check the links and <shhh> tell me if you want a copy of the pdfs </shhh>
12:39 bernie: booting again with updated kernel from the partition... and nope.
12:39 bernie: whatever. I'll just boot with external kernel and I'll come back to this mystery at some point in the future (in this iteration of the universe or the next)
12:41 bernie rgs__: hmm... not the type of books I'm reading these days... I'm mildly interested in ruby, but not so much
12:41 rgs__: btw, why isn't anyone using ruby 1.9 yet?
12:41 rgs__: is it a mismanaged release like python 3?
12:42 rgs__ bernie: nah, nah.. its an incremental release
12:42 bernie: no breakage. Its driven by Japs, they respect humans so much.
12:42 bernie rgs__: haha lol
12:42 rgs__ bernie: OTOH, Rails breaks API without giving a damn about developers. If its wrong it should be changed.
12:43 bernie rgs__: but then why do both fedora and ubuntu still ship 1.8.x after one year?
12:43 rgs__ bernie: western attitude of "your problem" if it breaks
12:43 bernie rgs__: this is typical of young and immature projects. it's also a necessity, at early stages
12:44 rgs__ bernie: errr.. now that you mentioned, i think they were a few minor incompats.. but the major objection against releasing it sooner (now that i think of it) might have been that performance tunning was still in place. Not sure, should look at the blogs
12:44 bernie rgs__: linux (the kernel) managed to maintain a very stable api since the beginning because it was based on solid standards such as POSIX and SUS
12:44 rgs__: early unices were continuously redefining the UNIX API...
12:44 rgs__: it's how evolution works!
12:46 rgs__ bernie: indeed
12:47 bernie: i think the passionate programming, which deals with software development issues, might be an enjoyable read for you. its only 200 pages
12:47 bernie rgs__: yes, it sounds nice. maybe send me the pdf
12:48 rgs__: I wonder what it says about turning down an offer from google
12:48 rgs__ bernie: haha
12:48 bernie rgs__: ah, I'm going to have lunch at google with morgan's boyfriend. but I'll refuse to sign their NDA. I wonder if they will admit me anyway.
12:51 rgs__ bernie: dude, it has *nothing* to do with the way the kernel is loaded (internal vs external). Its completely random!
12:51 bernie: sometimes it starts and sometimes it stalls for ever
12:51 bernie: maybe housetree is sick?
12:59 likes name.com
13:19 aldo <aldo!~Paraguay@mail.paraguayeduca.org> has joined #olpc-paraguay
13:19 aldo car|0s: buenas
13:20 car|0s aldo, buenas
13:22 bernie rgs__: it still seems related to initrd
13:23 rgs__: using the external kernel lets you specify the kernel command line
13:23 rgs__ bernie: nop
13:23 bernie rgs__: maybe add all the debug things you can possibly imagine
13:23 rgs__ bernie: it has nothing to do with the method
13:23 bernie: I've been able to boot with both
13:23 bernie rgs__: also try to upgrade the kernel in the vm to the latest version (and extract it to /srv/vm/kernel/...
13:23 rgs__ bernie: its just a matter of trying a zillion times
13:24 bernie rgs__: wow, that's odd
13:24 aldo has quit IRC
13:24 rgs__ bernie: its truly random, no kidding
13:24 bernie: but it makes sense.. because nothing from inside the VM was touched.. so something in the outside mas be broke
13:24 *must
13:26 bernie rgs__: I suspect initrd
13:26 rgs__: so try upgrading kernel, which will also remake initrd
13:26 rgs__ bernie: might be. I've also been able to boot with the new initrd
13:26 bernie rgs__: which new initrd?
13:27 rgs__ bernie: freshly re-created after aptitutde full-upgrade
13:27 bernie rgs__: (when I say initrd, I really mean initramfs... can't get used to the new naming)
13:27 rgs__ bernie: booting from the partition, as I told you I was able to do it.. after lots of perseverance
13:27 bernie rgs__: ah ok.
13:27 rgs__ bernie: has to be somethng with libvirt or housetree..
13:27 bernie rgs__: so I have no idea wtf is wrong with it. the other vms supposedly work
13:28 rgs__ bernie: supposedly. is not that they receive that much use that we would get complains right?
13:28 bernie: do we have sensitive clients on housetree?
13:29 bernie rgs__: I upgraded and restarted libvirt on housetree a few hours ago
13:29 rgs__ bernie: hmmm
13:30 bernie rgs__: we could reboot housetree if you want, but I doubt it would solve anything
13:30 rgs__: I tried starting the template vm, and that one started regularly
13:30 rgs__: maybe try a few times?
13:30 rgs__ bernie: could be
13:30 bernie: anyways, I'll be working on serendipity on the following days, so I'll eventually be very pissed of by the problem and I suppose then I'll have to fix it.
13:31 bernie: tend not to fix stuff until I am way pissed of.
13:31 *off
13:31 bernie: shit, forgot to take lunch
13:32 bernie rgs__: the secret of the good sysadmin is being proactive
13:33 rgs__ bernie: if you are not multi-tasking..
13:33 bernie rgs__: the most striking example of being proactive is having good backups :-)
13:33 rgs__ bernie: hey! did that one a week ago!
13:33 bernie rgs__: but also keeping machines up to  date, secure, etrc
13:34 rgs__ bernie: running maverick. checked.
13:34 bernie: "My life as a responsible Sysadmin". by Bernauuurdo Innocenti
13:34 bernie: "Cooking with Bernie" is much more of best-seller IMHO
13:35 bernie rgs__: lol
13:35 rgs__ bernie: would you worry about this message :
13:35 Oct 20 09:23:30 serendipity kernel: [   10.162115] JBD: barrier-based sync failed on vda1-8 - disabling barrier
13:35 JBD?
13:42 bernie rgs__: nope
13:42 rgs__: barriers often don't work on non plain disks
13:42 rgs__: they're broken also on most md setups
13:43 rgs__: jbd == journaling block device
13:43 rgs__ bernie: are they like checkpoints for journaled filesystems?
13:43 bernie rgs__: it's the ext[34] hournal
13:43 rgs__: barriers are the way that filesystems tell the block layer to flush previous requests to disk before proceeding
13:44 rgs__: it's been agreed upon by kernel  fs and block developers that barriers will die soon in the kernel and be replaced by a better technique
13:44 rgs__: search on lvm
13:44 rgs__: lwn
13:44 rgs__: gotta go, bye!
14:44 morganya <morganya!~chatzilla@> has joined #olpc-paraguay
14:47 morganya rgs__: just sent an update email! sorry I haven't had a lot of time for this in the last couple of weeks. things have been pretty crazy here.
14:48 brb
14:48 morganya has quit IRC
14:57 morganya <morganya!~chatzilla@> has joined #olpc-paraguay
16:07 morganya has quit IRC
16:18 car|0s has quit IRC
16:31 car|0s <car|0s!~carlos@karova.pol.una.py> has joined #olpc-paraguay
17:33 car|0s has quit IRC
17:53 car|0s <car|0s!~carlos@> has joined #olpc-paraguay
18:14 morganya <morganya!~chatzilla@> has joined #olpc-paraguay
18:34 morganya has quit IRC
18:43 morganya <morganya!~chatzilla@> has joined #olpc-paraguay
19:56 morganya has quit IRC
20:11 reubencaron_ has quit IRC
20:11 reubencaron <reubencaron!~reubencar@weka.laptop.org> has joined #olpc-paraguay
20:11 jasg has quit IRC
20:20 reubencaron has quit IRC
20:20 reubencaron <reubencaron!~reubencar@weka.laptop.org> has joined #olpc-paraguay
20:27 car|0s has quit IRC
20:29 jasg <jasg!~jorge@> has joined #olpc-paraguay
20:47 jasg has quit IRC
21:17 [cronos] has quit IRC
21:33 [cronos] <[cronos]!~scs@mail.paraguayeduca.org> has joined #olpc-paraguay
21:48 morganya <morganya!~chatzilla@> has joined #olpc-paraguay
21:53 tch has quit IRC
21:57 [cronos] has quit IRC
21:57 [cronos] <[cronos]!~scs@mail.paraguayeduca.org> has joined #olpc-paraguay
22:00 tch <tch!~tch@> has joined #olpc-paraguay
22:23 [cronos] has quit IRC
22:35 [cronos] <[cronos]!~scs@mail.paraguayeduca.org> has joined #olpc-paraguay
22:42 morganya_ <morganya_!~chatzilla@> has joined #olpc-paraguay
22:44 morganya has quit IRC
22:44 morganya_ is now known as morganya
23:02 walterbender <walterbender!~chatzilla@> has joined #olpc-paraguay
23:02 morganya has quit IRC
23:03 morganya <morganya!~chatzilla@> has joined #olpc-paraguay
23:10 morganya has quit IRC
23:13 morganya <morganya!~chatzilla@> has joined #olpc-paraguay
23:55 manusheel is now known as manusheel_afk

 « Previous day | Index | Today | Next day »     Channels | Search | Join

Powered by ilbot/Modified.