IRC Logs for #crux Wednesday, 2012-11-14

*** tj__ has joined #crux00:09
*** tj_ has quit IRC00:12
*** tj__ has quit IRC00:43
*** tj__ has joined #crux00:44
*** pshevtsov has joined #crux00:48
*** vaddi has joined #crux01:04
*** vee_ has joined #crux01:21
*** kInOzAwA has joined #crux01:27
*** kInOzAwA is now known as Guest5669701:27
*** vee__ has joined #crux01:29
vee__how on gods earth do you guys compile on atom processors. building the kernel took forever xD01:30
*** kInOzAwA___ has quit IRC01:30
*** vee_ has quit IRC01:30
prologicpatience :)01:40
vee__i got a kernel panic...all patience is gone. haha01:41
prologicIntel atoms kinda suck :)01:41
vee__eh. going to look for some chromiumos like distro to intstall on this thing01:42
prologicbut I have a server in the US east coast that runs crux 2.701:42
vee__a server? what exactly does it do?01:42
prologicruns stuff :)01:42
vee__ah. okay01:42
vee__servers usually tend to do that01:43
prologicdns, web, tunnelling, ssh, proxying01:43
prologicyou know the usual stuff: )01:43
prologicplan on running up crux 2.8 x86_64 shortly on another server of mine with web gate one installed and start offering free shell accounts01:44
vee__i really have to learn this whole server thing01:44
vee__im constantly asked to setup servers for businesses...dont know how01:44
vee__shame on me01:44
prologicit's just software really01:47
prologicno big deal01:47
vee__sure its fairly easy, just have to do it once01:47
*** vee__ has quit IRC01:54
pshevtsovjue, frinnst, Thank you! Now I have working git-svn. I had to rebuild git after installing subversion-perl and git-svn started working allright.01:56
frinnstwee01:57
*** pshevtsov has quit IRC03:41
*** pshevtsov has joined #crux03:43
frinnstdear god sbs2011 is an abomination03:56
*** kInOzAwA has joined #crux04:16
*** kInOzAwA is now known as Guest7939804:16
*** Guest56697 has quit IRC04:19
Romsterfrinnst, no worse than win 804:43
Romsterwell installing isn't the issue oit's how the f do you work it without a touch screen.04:43
Romsteri keep going for the menu04:43
Romsterno wonder the leadership changed because of that04:46
*** diverse has joined #crux04:47
diverseI am thinking of getting an ssd. How many of you guys have one and is it reliable?04:48
Romsteri was talking about one not long ago in #crux-devel04:48
Romsterhttp://crux.morpheus.net/irc/04:48
diverseI guess I should join the dev channel :P04:48
diverseok, so either Intel 330 or Samsung 83004:51
diverseI was thinking of installing crux on one of these.04:52
* diverse imagines the compile times on them04:52
diverseis there anything special about the controllers in any SSD?04:53
Romsteror wait for the newer ssd04:55
Romstercompile times nto so much as i compile in tmpfs already04:56
Romsterbut imagine the ccache performance compared to a raid0 setup04:56
Romstermost of the time is wasted on seeking04:56
frinnstthe controllers are generally what matters04:57
frinnstI have a 25-m, 310 40gb (i think) and two 33004:57
Romsteri might go for the 330 and migrate that to another boxen when the better ssds are out.04:58
frinnst*320 40gb04:59
diversefrinnst: which are the most reliable controllers?04:59
frinnstdunno about that. its usually more or less the same hardware in most ssd's05:00
frinnstintel usually supplies the nand, and intel themselves have first choice in the nand they use for their own05:00
Romsterhttp://www.news.com.au/national/mans-body-found-near-waterhole-in-mutawintja-national-park/story-fncynjr2-122651686046805:00
frinnst(i only buy intel)05:00
Romsteridiots in this country :/05:01
diversefrinnst: seems like intel is the only way to go. I have read many reviews about how users had their drives for at most 6 months before it died on them.05:02
Romsteri worry about that part data recovery, seems like a good idea to backup05:03
frinnstsamsung is probably a good choice too05:03
Romsteri run raid with everything else but running raid with ssd is not a good idea.05:03
frinnstand performance for general desktop usage is probably more or less the same05:03
Romsterkeep away form OCZ for bad service.05:04
frinnstyeah raid is never a valid backup solution anyways05:04
Romstertrue, but it covers availability when a disk dies.05:04
Romsteri need more hdd's for backing up too.05:05
diverseI heard bad things about OCZ in particular, but recently someone i know, praise about their product.05:05
frinnstyeah I have like 8 various drives in a btrfs array for backup :)05:05
Romsteryour trusting btefs alot05:06
* diverse still uses ext405:06
* Romster still uses ext305:06
frinnstwell its the only real option for that situiation unless i want to use zfs05:06
Romsteri guess so05:07
diverseRomster: why ext3?05:07
frinnstbesides, how big are the chances that both arrays will die at the same time? :D05:07
* frinnst imagines a bad btrfs commit to the linux kernel05:07
Romsterdiverse, ext4 didn't exist when i formatted and i have been too lazy to convert to ext405:08
diverseah05:08
Romsterplus i trust ext3 mroe than ext4 so far and the recent rare coruption bug in kernel didn't help my ability to start trusting ext4 being mature05:09
frinnstyet you run lvm? :D05:09
diversehmm, but I thought ext4 had mature quite a bit since its introduction, I had no problems so far.05:09
Romsteryeah crazy isn't it...05:09
diverseRomster: which kernel version?05:10
Romsterand mdadm and some kernels had that wipe metadata bug too.05:10
Romsterit's fixed now diverse05:10
frinnstyeah raid5 is dangerous without a ups05:10
diversefrinnst: go with raid1005:10
Romster3.6. something05:10
Romsteri'm on 3.6.8-pf05:11
diverse3.6.4 here05:11
Romsteri'm using raid1 multiple groups as each PV05:11
Romsteri should change that later on though05:11
Romsterbit silly using mdadm raid1 under lvm2 that can do mirror LVs too.05:12
Romsteri got sick of power outages so i got a UPS and a generator now.05:12
frinnstgenerator too? cool :)05:12
diverseRomster: I am a bit resistant on updating my Arch box, so I am stuck with that version until I switch to crux.05:12
Romsterall i need is some extra cables and rig up some relays for power changeover.05:13
Romsteryeah remote start generator :P05:13
Romsteronly a small 2.7KW generator but it'll be good enough to power 2 computers a fridge a freezer and a tv/radio05:14
Romsterdunno how long 20 liters of petrol will last though.05:14
Romsterguess we'll find out when the time arises.05:14
frinnst:)05:14
diverseI am also considering on getting a velociraptor as an alternative.05:15
Romsterfrinnst, it's a bit over the top but i get sick of the 5 plus hours without power when power sheading due to some transformer or far too many air cons are on a hot day.05:15
frinnstwith a full diesel tank i think we can power ourselves for about a month05:15
Romsteri'd love a diesel genset but this is just a petrol one.05:15
frinnstour system came with the location from the military05:16
Romsterso i'm gonna setup generator change over relay, ups on each pc.05:16
diverseWD came out with that 1TB drive recently05:16
Romsteri saw that diverse05:16
diverse*velociraptor drive05:16
* frinnst has had a velociraptor die once05:17
Romsterthough it's only 10rpm what ahppend to the 15krpm ones.05:17
diverse15krpms? Where?05:17
RomsterSAS disks can come in at 15Krpm beating a raptor but you need a SAS controller.05:18
Romsteri supose they don't want a SATA disk near the performance of a SAS disk.05:18
Romstervelociraptor are short stroke like a SAS disk is too.05:19
Romsterfrankly a SSD is a better choice though 1TiB might be useful in a velociraptor05:19
diverseI would like to go with an SSD if it were not for those reviews I read and worrying about losing data.05:20
diversewhen it comes to speed, yeah SSDs are awesome, no doubt. The slowest SSD is faster than the HDD raid 0 alternative. What I care about most is data integrity.05:22
Romsteri've hit one customers ssd that lsot all data no backup and a firmware bug made it not accessible until a reset. which wiped it.05:23
diverseSo it comes down to the firmware?05:24
Romsterand NAND quality i supose05:25
Romsterof course if the controller dies it's not accessible either.05:26
diverse;/05:26
Romsterwelcome to technology05:28
diversewelcome to *new* technology05:28
diverseHmm, they also make these PCI Express SSD embedded cards, which look interesting.05:30
Romstermpci05:31
Romstersome mobos have a socket for those.05:31
Romsterhttp://www.news.com.au/travel/holiday-ideas/salt-hotel-a-resort-for-all-season/story-e6frfqd9-122651645562305:33
Romsterdamn getting tired05:44
*** linXea has joined #crux05:55
pshevtsovHello! I did prt-get depinst skype but whent I try to start skype I get bash: /usr/bin/skype: No such file or directory06:03
pshevtsovI'm on x86_64 if it matters06:06
*** horrorSt1uck has joined #crux06:19
*** horrorStruck has quit IRC06:21
*** pshevtsov has quit IRC06:33
frinnsthah, just when i was about to respond06:33
*** pshevtsov has joined #crux06:34
frinnstwee06:34
frinnstpshevtsov: i was just about to say:06:35
frinnstsounds like you are trying to install a 32bit binary port06:35
frinnstyou get error messages like that when its linked against glibc 32bit06:35
pshevtsovfrinnst: So what should I do?06:36
frinnstthe skype port in contrib looks like a 32bit binary. You could get it to work with jaegers multilib06:36
frinnstyou can easily convert a pure 64bit crux to multilib06:36
frinnsthttp://jaeger.morpheus.net/linux/crux/files/multilib/06:37
frinnstit boils down to installing the glibc packages and gcc from that iso. and add the multilib repos to /etc/ports/06:39
frinnstsomeone else can give you more and perhaps correct details06:40
pshevtsovfrinnst: Sorry I don't get it. It isn't clear for me :(06:44
frinnstcrux x86_64 will not run 32bit binaries. You can install 32bit libraries to allow for 32bit programs (like skype) to work06:45
frinnstthis is called multilib. and the url i pasted contains these. You dont have to reinstall, but only replace a couple of packages from the ISO above06:46
frinnstI dont know exactly what you need to do so you could wait a bit for jaeger to wake up :)06:47
frinnstyou also need to make sure you have support for 32bit binaries in your kernel06:48
frinnstCONFIG_IA32_EMULATION=y06:48
*** horrorSt1uck is now known as horrorStruck06:50
juehmm, the skype port is rather outdated, there are 64-bit version available, I guess the version for ubuntu 10.04 should work without too much trouble07:01
*** horrorStruck has quit IRC07:01
jueI'd suggest to try that first07:02
*** joacim has quit IRC07:13
Romsterhttp://www.youtube.com/watch?v=1Cv2LqZwQUs07:25
*** joacim has joined #crux07:26
Romster64-bit guess i should look into that.07:26
pshevtsovHow to "extract" deb file?07:30
*** horrorStruck has joined #crux07:30
*** s44 has joined #crux07:32
*** spider44 has quit IRC07:36
pshevtsovWell, I've extracted skype 64-bit deb but inside is 32bit binary :(07:37
pshevtsovfile skype07:37
pshevtsovskype: ELF 32-bit LSB executable, Intel 80386, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.15, BuildID[sha1]=0xd6667f895e58d47966d6a572b13125e9110ee256, stripped07:37
juepshevtsov: sorry, didn't know that, what a nonsense :(07:39
*** kInOzAwA has joined #crux07:40
*** kInOzAwA is now known as Guest2211407:40
*** pshevtsov has quit IRC07:41
*** Guest79398 has quit IRC07:44
*** Kaishi has joined #crux07:51
Kaishihey all, I'm working on learning crux at the moment.  I just saw that version 2.8 is out.  I'm trying to update an older installation to support an in-production application.07:51
KaishiI saw the upgrade via CD instructions in the documentation, but the 2.8 release notes said it was best to not upgrade using ports because certain core utils and libs are ABI incompatible07:52
Kaishiwhat's the best way to do this?07:52
Kaishithe biggest thing is that I need the newer kernel provided in 2.8.  The environment in question is running in a VM and would benefit from the upgraded VM support in more recent linux versions07:55
Kaishiprt-get sysup is as much as I've done so far, but a good number of packages have failed that, probably because of that ABI incompatibility07:55
*** tj__ has quit IRC07:58
*** tj_ has joined #crux07:59
*** dkoby has joined #crux08:11
*** Guest22114 is now known as kInOzAwA08:18
*** kInOzAwA has joined #crux08:18
jaegerKaishi: without specific errors we probably can't help you very much, which things are broken?08:21
Kaishithanks for the reply :) I'm not exactly sure! I'm still learning crux08:21
Kaishibut, prt-get diff shows a lot less differences now08:22
KaishiI've just been running prt-get sysup followed by a reboot, followed by prt-get diff08:22
Kaishirepeat08:22
Kaishiand after 2 passes, I'm down to a small number of packages that won't update08:22
jaegerYou might also want to run revdep from the prt-utils port, there are likely some linking issues to be sorted08:22
jaegerKaishi: Though it isn't the ideal way to upgrade, if you need to do it on a running machine you can mount the ISO and pkgadd packages from it, just for future reference08:26
Kaishihow would I invoke that?08:26
KaishiI mostly just want the new kernel08:26
jaegerwell, you should be able to install a new kernel almost any time :)08:26
jaegerinvoke which, revdep or the pkgadd upgrades?08:27
Kaishirevdep08:27
horrorStruckKaishi: libpmc update breaks gcc08:27
horrorStrucklibmpc*08:27
Kaishithings still seem to be working at the moment, how would I know if I broke it, and how would I fix it again??08:28
jaegeryou can run it by itself to get a list of ports that have linking issues. if you want more info, try RD_VERBOSE=2 revdep08:28
Kaishijaeger: so just revdep ? I tried that already08:28
Kaishialso prt-utils revdep08:29
jaegerjust revdep08:29
jaegerdid you install the prt-utils port?08:29
Kaishithe crux install I have to use is about a year and a half old >_<08:29
Kaishiand I don't have the option to reinstall right now08:29
Kaishibecause of the in-production application it is hosting08:29
horrorStruckKaishi: what does this return? readelf -d /usr/lib/gcc/i686-pc-linux-gnu/*/cc1|grep mpc08:32
Kaishi"0x0000001 (NEEDED)   Shared Library: [libmpc.so.2]08:34
horrorStruckand this: prt-get current libmpc08:34
Kaishihad to type the input and output again because I can't copy-paste into or out of the VM connection, without that updated kernel >_<08:34
Kaishi0.9-108:34
horrorStruckyour gcc is broken08:35
Kaishiyour name is appropriate for my emotions08:35
Kaishi>_<08:35
Kaishiso how do I unbreak it?08:35
horrorStruck:D08:35
horrorStruckrebuild it08:35
Kaishire..build...gcc?08:35
horrorStruckor get from the 2.8 iso08:35
jaegerI'd suggest readding it from the ISO08:36
jaegersaves a lot of time and effort, especially if the current one is broken08:36
Kaishisure, I can do that, I have the ISO and I can mount it shortly.08:36
Kaishi I'll need a few minutes, have to afk08:36
jaegerBefore you do that, I'm curious what state your system is in. You said you've done some port updates but what you really wanted is just the kernel... did you change your /etc/ports/* to point to 2.8 and now you're somewhere between 2.8 and whatever version you had running?08:37
horrorStruckwait08:37
horrorStrucki'm wrong08:37
Kaishijaeger: I changed almost nothing, and it had been minimally updated since installation08:38
horrorStruck0.9 does provide libmpc.so.208:38
horrorStrucksorry08:38
jaegerok, what DID you change?08:38
Kaishiit was originally brought up a while back to host a basic web application, which became increasingly important to our team08:38
KaishiI just ran prt-get sysup and rebooted it a few times08:38
KaishiI haven't done anything manual08:39
jaegerok, so adding things from the 2.8 ISO is probably NOT a good idea, then08:39
Kaishino changing mount points or configs08:39
Kaishican I check the crux version easily? I don't even know how old this really is08:39
KaishiI was told about a year and a half08:39
horrorStruckcrux08:39
Kaishiah, it's 2.708:39
Kaishiso that's not too far08:39
Kaishiright?08:40
KaishiI really do need to step away, can you guys give me like 15 minutes?08:40
Kaishihave to talk to my boss ^_^08:40
jaegermost likely will be here08:40
Kaishicool, thanks :)08:41
horrorStruckprotip: blame someone else08:42
cruxbot[opt.git/2.8]: lm_sensors: updated to 3.3.308:48
cruxbot[opt.git/2.8]: gnutls: updated to 3.1.408:48
cruxbot[opt.git/2.8]: cmake: updated to 2.8.1008:48
Romsterhttp://i.imgur.com/eSprT.jpg09:04
KaishiI'm back, sorry about that09:04
*** sepen has joined #crux09:05
sepenhi09:05
Romsterhi sepen09:05
Kaishijaeger: so if I'm on crux 2.7 and have partially updated the ports, what should I do now?09:05
jaegerKaishi: is anything actually wrong with the system right now? I don't understand what you're trying to do anymore.09:06
jaegerheyo, sepen09:06
horrorStruckhi sepen09:06
Romsterthere have been a few changes since 2.7 was first out...09:06
Romstermodule-init-tools removed added kmod... something else gcc related i can't quite remember.09:07
Romstera sysup on it's own would break past that point.09:07
Romsteroh that dependency for udev and devtmpfs support in kernel needs added too.09:08
KaishiI'm wondering if it's even safe to reboot right now09:09
Romsterdid you update udev?09:09
Kaishithings seem okay at the moment, but when I hear "broken" I think, "keep it alive and rescue without giving it a chance to fail even worse"09:09
Kaishithe only thing I have done is a couple of sysup09:10
Kaishibut with reboots between them09:10
Kaishinothing forceful09:10
Romsterand what did that update and what is not building?09:10
Kaishia bunch of things, the overwhelming majority of the packages in fact.  There are maybe 10 or 15 that aren't current at the moment09:11
Romstercore ones?09:11
Romsteri would download this http://jaeger.morpheus.net/linux/crux/files/updated-crux-iso/ mount the iso somewhere and cd to core on the iso09:12
Romsterand do: pkgrm module-init-tools ; for p in *; do pkgadd -u $p; pkgadd $p; done09:12
jaegerI'd much rather know the cause of the problem than start reinstalling packages to work around something09:13
Romstermake sure you recompile kenrel with devtmpfs and automout dev tmp fs09:13
Romsterand prt-get update `revdep`09:13
jaegerKaishi: if your immediate concern is whether or not the machine will boot properly, and assuming you DID update udev and kmod, you need these in your kernel config:09:14
jaegerCONFIG_UEVENT_HELPER_PATH=""09:14
jaegerCONFIG_DEVTMPFS=y09:14
jaegerCONFIG_DEVTMPFS_MOUNT=y09:14
Romsterwell i suppose Kaishi could test that gcc is working still for starters if jaeger feels like a debug session.09:14
cruxbot[contrib.git/2.8]: vlc: 2.0.3 -> 2.0.409:14
cruxbot[contrib.git/2.8]: subversion-perl: updated to 1.7.709:14
cruxbot[contrib.git/2.8]: subversion-python: updated to 1.7.709:14
Kaishijaeger: the only problem I have encountered thusfar is the outdated kernel.  Beyond that, things are working as needed, just old.09:14
jaegerIt's my personal opinion that finding the root cause of a problem and fixing it is *always* the better choice. That's just opinion, though, not everyone agrees09:15
KaishiI'm not panicked yet09:15
jaegerKaishi: ok, that wasn't clear to me since you've mentioned sysups several times09:15
Kaishithe problem right now is, "this is old and I don't trust it."09:15
KaishiI know I could just keep running it as it was, but the kernel won't support the clock sync from the VM host09:16
Kaishiso time will slide out of sync09:16
Kaishiand I could fix using NTP or something09:16
Romstereven with old kernel your sysup should in theory work for all ports.09:16
Kaishiromster: that's what I"m thinking09:16
KaishiI didn't invoke anything beyond that09:16
Romsterbut i don't kwo about if it'll come back up after reboot, probably not.09:16
Romsterknow*09:16
Romsterso pastebin logs of ports that failed?09:17
Kaishiyou don't think it'll come back after a reboot? I could take a snapshot and try rebooting it, roll back if it fails09:17
Romsternot if you bumped udev.09:17
KaishiI can't copy-paste any files in or out of the system at the moment, my only external access is via a web interface09:17
Romsterif it's older udev then it should probably boot ok.09:17
Romsterwgetpaste?09:18
Romsterin contrib09:18
jaegerIf you did update udev, just make sure you have those kernel config options listed above09:18
KaishiI mean it's a fully commandline environment right now, bash only09:18
Kaishiudev is 162-109:18
Kaishiis that old or new?09:19
jaegerolder09:19
Romsterok sounds like the old one.09:19
jaeger182-2 is current09:19
Romsteri think that will work without them other options.09:19
Romsterin kernel .config09:19
KaishiI wouldn't think that running a basic sysup would break anything09:19
*** dkoby has quit IRC09:19
Romstersome cases ports change and manual intervention is needed. when that happens a mailing list post is sent out.09:20
jaegerIn most cases it doesn't, but every now and then it happens09:20
KaishiAh.09:20
KaishiWell, I snapshotted it in the immediate state, I'm going to try rebooting it, and either keep that or roll-back if it doesn't work09:20
Romsteroh i was thinking of libeffi for glib but that got moved to opt again so disregard that one.09:21
Kaishi"warning, the state track is obsolete, use conntrack instead"09:21
Kaishiis the only thing I saw on boot that was ... odd I suppose09:22
Kaishibut it works, or seems to09:22
Romsterjust a warning it'll still work, there is conntrack in contrib.09:22
jaegerthat message probably came from iptables09:23
Kaishiah09:23
Romsterat some point that warning will become fatal, but hasn't happened yet.09:23
KaishiWhat should I do about it? Pick up conntrack?09:23
jaegerconvert your state match rules to conntrack rules09:23
Romsteriptables is depreciating conntrack in the iptables port.09:24
KaishiI see09:24
Kaishihow would I check those rules? It might be a very basic config that would be replaced with a new default config09:24
Romsteras the majority will be on IPv6 and so it's delegated to it's own conntrack port.. and marked as obsolete in the iptables port.09:24
jaegerit should be pretty simple like changing "-m state --state RELATED" to "-m conntrack --ctstate RELATED"09:25
KaishiI don't even know where those rules would live09:25
Romster/etc/rc.local perhaps09:25
jaegeryou can list the rules in place currently with "iptables -L" but you'd need to find where they're set. /etc/rc.d/iptables or /etc/rc.local or something, maybe09:25
Romsterbest approach is grep for iptables-save in /etc/09:26
Romstersee if it's saved anywhere and iptables-restore on start up.09:27
Romsteror you many have a ton of iptables rules stashed somewhere in /etc/ grep will find them.09:28
Kaishithere are only like 8 rules :)09:28
Kaishilooks like it's all bascially wide open, with an explicit rule created for the application we're running09:28
Romsternot everyone needs tons of rules09:28
KaishiWe have some pfsense boxes between us and the internet so it'll be fine09:29
Romsterdid you set this VM up a long time ago or picked up from someone else?09:29
Kaishibut I am generally in favor of host-based firewalling to be sure09:29
Kaishisorta a team project, but I didn't do the initial setup09:29
Kaishi /etc/rc.d/iptables doesn't exist so not htere09:31
Kaishistill looking09:31
jaegerif that doesn't exist it's likely that iptables-save is being called somewhere09:31
Kaishijust checked /etc/rc.local09:32
jaegerit could be in /etc/rc.local, it could be in /etc/rc.d/net, it could be somewhere else09:32
KaishiI'm seeing iptables-restore09:32
Romsterkernel compilation is done manually i usually do wget ....linux-3.6.4.tar.bz2 ; tar xjf linux-3.6.4.tar.bz2 ; cd linux-3.6.4 ; zcat /proc/config.gz ./.config ; make silentoldconfig ; make menuconfig09:32
Romsteredit them options that udev needs.09:32
jaegerIt could also be that iptables-save was only run manually by someone logged in, then iptables-restore is run to enable those rules when booting09:32
Romsterbtw cd /usr/src/ for the spot kernel sits in.09:32
jaegerOn a side note, pfSense is great09:33
Romsterrest is the usual make and copy System.map and bzImage to /boot issue lilo if using lilo.09:33
Romsteras for compilaion failures i know that you can't update udev until you pkgrm module-init-tools and prt-get install kmod09:34
Romsterthen prt-get update udev09:35
Romsterbut don't do that until you got a kernel with them options jaeger mentioned earlier.09:35
Romsterrest of the failures we would need to see those.09:35
Romsterprt-get update -fr `revdep`09:36
Romsterwould be an idea too.09:36
Romsteras previously mentioned.09:36
Kaishiso rather than manually compiling all the updates, would it be easier / safer / faster to toss in the 2.8 ISO, boot to that installer, and follow the update procedure?09:36
Kaishishould I fix the iptables thing before I do that, if so?09:36
Romsteri wouldn't goto 2.8 just yet without testing it wont break anything important. but that's my opinion.09:37
Kaishiromster: I generally agree, but I have a snapshot in a safe state, so that's the test.09:37
KaishiI can always roll it back09:37
jaegerI would ONLY upgrade to 2.8 after you test that your web application or whatever else works there09:37
jaegersince you're already doing this in a VM, you could clone the VM and update the clone for testing purposes09:37
jaegerthat way you can do whatever you like to the clone without affecting any production uptime09:38
Romstergood idea.09:38
KaishiI could clone it, but that's what snapshots are for09:39
KaishiI have a clone on disk, of t he whole state, including the HDD and the RAM state09:39
jaegerwe'll have to agree to disagree on that one, heh09:39
Kaishiwould take about 10 second to roll back09:39
jaegerIt's up to you, of course, it's your environment. That was only my suggestion.09:40
Kaishino I understand, if the problems encountered aren't immediately apparent, then keeping a safe clone for a few days/weeks is a good call09:40
Kaishibut, I can just keep the snapshot that long for the same effect09:40
Kaishiminor performance penalty but the host has hardware to spare09:41
jaegerYes, my concern is that any time you do a rollback, state is lost09:41
jaegerFor example, if you're logged in via SSH and you do a rollback, that session gets disconnected09:41
jaegerthat sort of thing09:41
Kaishithe only connections are via the local terminal or the web server09:41
Kaishino SSH :)09:42
Kaishibut great point09:42
KaishiSSH comes later, hasn't even been configured, that's how basic this environment is09:42
jaegerIn your environment perhaps it isn't a valid concern, at least in this case, but that's what I was getting at09:42
Kaishimhmm! it's certainly valid09:44
Kaishithe main usage hours for this are in the afternoon / evening, but it's safe for me to do whatever I need to right now09:44
Kaishibut I'd like to have it done by 2 PM EST, ideally09:44
Kaishiwhich shouldn't be impossible09:44
Kaishithe ports updates are all well and good, but I can't update the JRE or the kernel, which are the two biggest performance limiters for this particular config09:45
Kaishiso I'm thinking if I go to 2.8, it'll update both of those, most likely09:45
jaegerWhy can't you update them?09:45
KaishiI'm not sure, but the JRE update failed and I didn't pry09:46
KaishiJRE is currently 7-linux-i58609:46
Kaishiso vanilla 1.7 I'd assume09:47
Kaishibut that's almost a year old I think09:47
Kaishimaybe more09:47
jaegerI think 7u9 is current09:47
Kaishicorrect09:47
jaegerwith that said, CRUX 2.7 is now superseded by 2.8 so 2.7 won't be getting all the updates anymore09:47
Kaishimhmm09:48
jaegerwith THAT said, you could still override the port with a port of your own, or install the JRE manually09:48
Kaishiokay, the error is in building the JRE09:48
Kaishilooks like an md5sum mismatch09:48
Kaishiin the jre-7u9....tar.gz09:49
jaegerIf the port expects a different version than the current one that's to be expected09:50
jaegersince the downloads won't be the same file09:50
jaegerYou could override the port with an updated one or download the JRE version that the port is expecting09:51
Kaishicould I force it to take a new copy of the source from the ports collection?09:52
Kaishiwould that be safe?09:52
jaegerwell, the JRE source archive won't come from ports, it's a bit of a special case in that you need to agree to oracle's license agreement before downloading it09:52
jaegerIf you need jre-7u9 I would suggest you make a copy of the jre port, update the copy so that it expects 7u9, and install that09:53
Kaishiso what's being hosted on the ports collection?09:53
KaishiIf that's just a binary or something, then that's fine, I don't really care either way09:54
jaegerjust the files that tell the system how to install the jre09:54
KaishiI see!09:54
jaegeryou download the jre manually and the port tells crux what to do with it09:54
KaishiSo, how would I override this md5sum mismatch and just get a new copy of the installer from prt-get?09:54
Kaishihmm09:54
KaishiI don't really want to download anything manually if I can help it09:55
jaegerin the case of JRE you have to. most things you don't09:55
jaegeralso, it appears that the jre port in CRUX 2.7 is already updated to 7u909:56
jaegerSo no override is needed09:57
Kaishiprt-get update jre yeilds the following:09:57
Kaishi===> Error: Md5sum Mismatch found:09:58
KaishiMISSING  56178.... jre-7up-linux-i586.tar.gz09:58
KaishiNEW    d41d8.... jre-7u9-linux-i586.tar.gz09:58
jaegerthe md5sum should be 56178ed00dab2ebd8268caf5575743f4  jre-7u9-linux-i586.tar.gz09:59
Kaishi===> Error: Building '/usr/ports/opt/jre/jre#1.7.0_09-1.pkg.tar.gz' failed.09:59
jaegerif yours doesn't match that, delete the file and re-download it09:59
KaishiHow would I delete it? Where is it?09:59
jaegerwhat does "grep PKGMK_SOURCE_DIR /etc/pkgmk.conf" say?10:00
jaegerthat will tell where source files should go10:00
Kaishi="$PWD"10:02
jaegerso it should be in the port's directory, /usr/ports/opt/jre10:02
*** sepen has quit IRC10:38
Kaishijaeger: I got my RDP session disconnected for a bit, but I'm back10:39
Kaishiso should I just rm jre-7u9* ?10:39
Kaishior something akin to that?10:39
Kaishiheh, I did that and um10:40
Kaishinow the prt-get operation fails a whole different way10:40
joacimif i'm not mistaken, you have to download the file manually10:43
joaciminstructions should be in the readme10:43
jaegeryes, as joacim says.10:44
jaegerrm the old file, download the new one10:44
Kaishiokay, I'd love to do just that, but I'm doing this entirely from command line10:45
Kaishiam I going to need lynx or something?10:45
jaegereither that or some other way to get the file to the VM - scp, ftp, http, usb, etc.10:46
KaishiI have the file on my windows workstation10:48
Kaishibut I don't think crux has samba or anything like that installed.10:48
KaishiI could check SCP, maybe that would work10:48
Kaishiaccess denied >_>10:50
jaegerscp requires an ssh server which isn't in windows by default10:50
KaishiI'm using WinSCP10:50
jaegerthen the server would need to be allowing ssh connections10:50
KaishiI really need to make myself my own user account10:51
Kaishihmm10:51
KaishiI am such a noob at this10:51
KaishiI feel really dumb10:51
Kaishiuseradd, that's right10:55
Kaishiand then passwd10:55
Kaishigotta change groups though >_> hmm10:55
Kaishiokay, useradd -G10:58
Kaishiso is there a sudoers file or group or something?10:58
Kaishihas rights but isn't always elevated?10:59
jaegerby default there's not a specific group defined, though someone may have done so on your install if you're not the only one managing it11:00
jaegeruse 'visudo' to edit the sudoers file11:00
Kaishihmm11:02
Kaishiso where in the file do I add it?11:03
*** tj_ has quit IRC11:03
KaishiUser_Alias?11:03
Kaishiohh, maybe user privilege specification11:04
Kaishiokay, permission denied to copy the JRE .tar.gz into the right folder from winscp11:06
KaishiI made my user account a member of root, at least temporarily, or at least I attempted to11:07
Kaishido I need to chmod / chown the folder before I can copy into it?11:08
Kaishilooks like I'm not in root or something11:10
jaegerjust copy it into your user's directory, then move it from the server's command line11:13
Kaishiyeah, htat's what I just did11:17
Kaishithat makes a lot more sense11:17
KaishiI was being lazy11:17
Kaishi>_<11:17
*** tj_ has joined #crux11:19
Kaishiwoop, the jre is installing11:20
KaishiI'm pretty happy about that11:20
Kaishiawesome11:21
Kaishiall done11:21
Kaishihow do I check the kernel version?11:21
Kaishiuname -r I suppose11:24
KaishiI'm seeing 3.1.0-rc911:24
Kaishithat's really old11:24
KaishiI need to be in the 3.4+ to have the drivers I need for the VM host11:24
Kaishiapparently we had the JRE in a non-standard location12:00
Kaishiand upgrading to the latest JRE from oracle has moved it to the standard location or something12:00
Kaishibut I can't find it anymore12:00
Kaishior maybe the symbolic link is missing12:02
jaegertry pkginfo -l jre12:03
niklaswehmm12:06
* niklaswe thinking his nfs time out, when he trying to mount it on my xbmc over internet..12:06
Kaishiwe were calling java from:12:29
Kaishi /usr/java/java-latest12:29
Kaishibut it looks like oracle puts the symbolic link in: /usr/lib/java12:29
Kaishiwhich just points to the appropriate JRE version, in /usr/lib12:29
KaishiI could try to find all the references to /usr/java/java-latest and change them12:29
Kaishior, I could just make an extra symbolic link12:30
Kaishihmm12:38
KaishiI'm having a lot of trouble actually bringing the JVM back online now12:38
KaishiI may need to roll back my VM just because the latest jre and our scripts don't match12:39
*** mike_k has joined #crux12:50
*** Rotwang has joined #crux12:51
KaishiI'm contacting one of our developers for his thoughts, but for the evening I'm just going to roll back to that saved state I made.12:52
jaegerok12:53
Kaishibut I'm not giving up!12:53
Kaishiso much more to try, I just have to have the application working for the core hours, which start in about 5 minutes12:53
jaegerThis is where I'd suggest cloning the machine again :)12:53
jaegerThat way you can do whatever you want on the clone while the live one stays live12:54
Kaishifirewall rules, mac addies, hostnames, all these things would need to change12:57
Kaishinot worth doing all that12:57
* jaeger shrugs12:57
Kaishino worries mate :) I understand your reasoning but it won't work for me in this case12:58
KaishiI'm down to ~10 packages that I can't update without troubles12:58
Kaishiand I might be able to reduce that number even further12:58
Kaishibut I know to leave udev alone12:59
Kaishiand glib|c right?12:59
jaegerI wouldn't say you NEED to leave any of them alone, as long as you're aware of what you might need to add/update12:59
jaegerthe udev stuff we've already gone over13:00
Kaishiyeah, don't update one part of that chain unless you can update all the others13:00
jaegeras for glibc it's usually not a big deal unless you're moving to a new major release13:00
Rotwanghi13:01
Kaishitried to update cmake, "Could NOT find EXPAT"13:03
*** lasso has joined #crux13:20
*** mike_k has quit IRC13:30
frinnstholy shit that's one hell of a backlog13:46
niklaswe=)13:56
Kaishiokay, time is crazy out of sync right now13:57
Kaishihow do I sync time in crux from command line?13:57
niklasweKaishi: set it with date13:58
niklasweor use ntpdate <ntp.server.com>13:58
KaishiI don't have ntpdate13:58
KaishiI don't see a package named ntpdate13:59
niklasweKaishi: it´s in pcakges ntp14:00
frinnstjaeger: you should send him a bill :)14:00
frinnst5 hrs :)14:01
jaegerheh14:02
jaegerKaishi: try rdate -n <ntpserver>14:02
Kaishicurrently compiling package ntp :<14:02
Kaishiif this should fail then yes14:03
jaegerrdate should already be installed but you can also use ntp if you prefer14:03
KaishiI don't prefer anything, I didn't know about rdate14:03
Kaishiwell alrighty, that's working a lot better now14:03
KaishiI'd really like to put a cron job in place to sync time pretty often14:04
*** joacim has quit IRC14:04
Kaishibecause it's a VM without any real timekeeping method at this point14:04
*** joacim has joined #crux14:04
Kaishiit's just... going.  It'll become more and more off over time, until I upgrade that kernel14:04
jaegergah... no matter how many times I ask for them, people keep sending in tickets with no error messages14:06
jaegerI do NOT read minds! I CANNOT see your computer screen at all times! You MUST give me some useful information!14:06
jaegersomeone tells me they have a problem with their .tcshrc file, I ask for error messages. they send me standard output from a program that isn't even related in any way to tcsh14:07
jaegerand contains no errors14:07
niklaswejaeger: heh14:08
joacimthey keep paraphrasing the error when i ask for logs or program output14:11
jaegerI'm slowly getting closer to the issue, which I believe is this: the user is running bash, not tcsh14:16
Kaishicron is fixed14:17
Kaishierm14:17
Kaishicron job to keep time in sync is in place, hourly.14:17
jaegerI had to do that in the past on a set of servers that had terrible RTCs14:22
Kaishiif the hvtime module were in the kernel it would work fine14:24
Kaishibut it wasn't officially added until like 3.414:24
Kaishigrr... looks like some other package related to scripts or something has broken14:27
Kaishinot sure which or why14:27
Kaishicertain webapp functions aren't giving output even if they work so I'm guessing PHP14:27
jaegerI win. user was running bash, not tcsh14:30
Kaishiall the backend scripts are python14:30
jaegercheck the webapp's error logs?14:30
Kaishilooking for such a thing >_> where are logs normally put?14:31
jaegerdepends entirely on the app14:32
jaegeror the app server14:32
Kaishihiawatha is the webserver14:32
KaishiI think14:32
jaegerlook in /var/log, if there's nothing there then check the config files for log locations14:33
Kaishiis there a way to reset the terminal font?14:38
KaishiI ran cat on a log and now my font is all busted.14:38
Kaishibut on tty2-6 it's fine14:38
jaeger'reset'14:43
Kaishiexcellent14:43
KaishiI considered just trying things like that14:43
Kaishibut I didn't wanna break anything so I went to tty214:43
*** hcaulfield57 has joined #crux14:44
*** hcaulfield57 has left #crux14:48
*** vee_ has joined #crux14:54
*** Kaishi has quit IRC15:18
*** kInOzAwA_ has joined #crux15:24
*** kInOzAwA has quit IRC15:27
*** sh4rm4 has quit IRC15:42
*** Kaishi has joined #crux15:42
*** sh4rm4 has joined #crux15:45
*** kInOzAwA__ has joined #crux15:48
niklaswegood evening15:50
frinnst4-215:51
frinnstgood night :)15:51
*** kInOzAwA_ has quit IRC15:51
niklaswe:)15:51
niklaswegood nighty then frinnst :)15:52
niklaswe http://i.imgur.com/JASkW.jpg =)15:56
*** Rotwang has quit IRC15:57
joacimthat is why i dont work15:58
frinnsthttp://www.youtube.com/watch?v=EQhuE3b_GBU&noredirect=115:58
joacimwhat show/movie is that from?15:58
niklaswe:)15:59
niklaswehttp://imgur.com/gallery/Nbweo16:00
frinnstoffice space16:01
joacimyou know, that panel about 9gag is what users on 4chan say about reddit :]16:03
*** lasso has quit IRC16:05
*** n0nsense has joined #crux16:29
*** n0nsense has joined #crux16:29
*** n0nsense has joined #crux16:29
*** n0nsense has joined #crux16:29
tilmanniklaswe: the doc's answer is even better16:43
tilman"wow. that's messed up."16:43
*** joe9 has joined #crux16:53
*** vaddi has quit IRC17:00
*** joacim has quit IRC17:04
*** joacim has joined #crux17:06
KaishiHow do I compile a new kernel? I'm already chroot'd in17:13
Kaishibut I don't know how to proceed17:14
Kaishithe Handbook just says "configure and compile new kernel!"17:14
Kaishilike that's the whole thing17:14
joacimsource-based distros generally expect that you already know how to compile your own kernel.17:20
joacimi think crux installs some sources in /usr/src by default17:21
Kaishiit's been probably 5 years since I've had to do that, since most major distros have gone to binary-based package distribution17:22
joacimhttp://crux.nu/Wiki/KernelBuilding17:22
KaishiI already found some documentation on it17:22
KaishiI really just need the Microsoft Hyper-V client drivers to be included.  Could have probably skipped upgrading any other packages and only upgraded the kernel, which would have been smart but the damage is done17:23
*** vee_ has quit IRC17:25
*** joacim has quit IRC18:00
*** joacim has joined #crux18:01
Kaishiafter all that, kernel panic18:09
Kaishiawesome :)18:09
Kaishisounds like we might have to just restore the database and rebuild some scripts on a newer install18:27
*** kInOzAwA___ has joined #crux19:43
*** kInOzAwA__ has quit IRC19:46
Kaishilooks like my kernel doesn't have "proc support" so I can't use zcat for kernel config19:53
*** tilman has quit IRC20:25
*** tilman has joined #crux20:28
*** ChanServ sets mode: +o tilman20:28
*** __mavrick61 has quit IRC21:37
*** joe9 has quit IRC21:37
*** _mavrick61 has joined #crux21:38
*** nthwyatt has quit IRC23:07
*** nthwyatt has joined #crux23:14

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!