| |
| |
| |
|
Page: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
Comments:
<0> dli_ jfs is usually recommended to delete a few number of large files, not a large number of small files
<1> i am going to run dd on part and see what some of the data looks like
<0> blackace aye, but with a normal myth box you could be recording and deleting 100s of GB's per hour and expect it to happen without slowing down the video being displayed
<2> thehailo, yes, that's what I found, but reiser4 is great with large number of small files
<1> aha, it is saying generated by LVM2
<0> jubo1 did you try 'reiser' or 'reiserfs' ?
<3> thehailo: isn't XFS good for that kind of thing too?
<0> dli_ aye, but I'm not touching reiser4 yet
<4> anyone know why azureus pulls in avalon logkit 1.2 when 2.0 is not masked and installed? 1.2 fails with "target 1.2 does not match source 1.5" which the #java guys seam to think is a problem with the ebuild
<0> LSD` I love and use XFS for everything, but every system I have is on a UPC
<0> *UPS
<1> thehailo: I tried both
<5> jubo1: if you have the lvm2 tools handy try "vgscan" and "vgchange -a y" (you may need to fiddle with lvm.conf beforehnad)
<2> thehailo, reiser4 saves space with small files too, qsize -fs shows 5-10% less space used than reiserfs
<3> thehailo: XFS never caused me any grief without a UPS, even on a box that my folks would keep switching off
<6> While trying to emerge mercurial, I get this error: I/O error : Attempt to load network entity http://docbook.sourceforge.net/release/xsl/current/manpages/docbook.xsl ; where could I find this file somewhere else, and where should I put it in my gentoo to be able to emerge mercurial?
<0> jubo1 if they didn't use ext2, ext3, reiserfs, jfs, or xfs, then I dunno what to say man lol
<3> thehailo: There was one incident with 2.4.22 but that kernel was a mess of problems I've since found out
<1> thehailo: well I will play with the lvm2 as kerframil suggested above
<0> LSD` ya personally I've never had issues, but I'm weary to just say "sure go use XFS on your database server with no backup battery" because of horror stories I've heard
<0> LSD` but between pushing Gbps of throughput and allowing for 9EB of data...I'll be sticking with XFS for awhile
<6> Why do I get "I/O error : Attempt to load network entity http://docbook.sourceforge.net/release/xsl/current/manpages/docbook.xsl"; error while emerging mercurial, but I can wget it?
<2> pjb, I suppose you save it to /usr/portage/distfiles
<6> Well, I'm not so sure, it seems to be downloaded by xml tools while compiling the mercurial documentation...
<7> morning...
<1> kerframil: ok so lvmdiskscan says it is a physical volume
<5> jubo1: then it's type should be 8e ... grrr
<5> jubo1: in any case, you should be able to activate the volume group to which the PV(s) belong in the manner I've alreay described
<1> kerframil: hmm, so did the knoppmyth installer forget to set that properly?
<1> kerframil: i remember something similiar holding me back on software raid. ahh memories...
<6> It's xmlto man ${x}.xml which cannot download the docbook.xsl
<5> jubo1: it's academic I believe, but it sure would have made it easier to identify ;)
<8> Gentoo looks like an old school mac...
<5> jubo1: with software raid the type has significance in terms of the kernel being able to auto-***emble arrays
<5> jubo1: with lvm, that isn't the case
<1> kerframil: yeah... btw, once I mounted it with lvm it works!!!!
<5> jubo1: excellent
<5> jubo1: and what filesystem as it in the end?
<5> was*
<9> is it possible to make multiple users own a file?
<10> i have 1.5GB of ram right.. and im running vmware with 600MB and many other Mem killing apps, how come gkrellm reports me only useing a tiny fragment of my total memory, whereas ksys says im using like 99% of my memory.. explain!! please;)
<1> kerframil: ext3
<5> signature16: use POSIX ACL support or make hard links to the file and set different permissions on each link perhaps
<11> signature16: put them in a common group
<5> jubo1: heh
<10> signature16, yeah its called a group
<1> kerframil and thehailo thanks a lot for the help
<1> my friend is literally in tears beside me
<5> jubo1: ooo er
<11> hahah
<9> brianw: so if i want root and userx to own /home.....would i jsut maybe put them both in wheel and set wheel to own /home?
<11> don't have to worry about root
<10> signature16, root owns all
<5> signature16: no. a group cannot be an owner.
<12> Anyone know a good program to catalog @50,000+ files, non GUI
<5> and yes, root owns all unless some sort of role-based access control mechanism is in place
<10> can anyone explain my crazy memory problem
<9> kerfamil: chown wheel: /home doesnt work?
<12> signature16: Don't you mean chown :wheel /home ?
<13> [02:49:07] <@kerframil> signature16: no. a group cannot be an owner.
<5> signature16: as I just said, a group cannot be an owner
<5> signature16: a group can be the ***ociated group :)
<5> Garibaldi: ah, thanks
<13> kerframil: np
<9> kerframil: so how do i have multiple users have complete contorl over /home?
<11> signature16: in your situation, there is no need, root does not have to own it
<5> signature16: describe exactly what rights you want to delegate and to whom
<5> signature16: I'm not clear on what you mean by complete control over /home (not sure why you'd need to change the perms on that paticular dir)
<9> kerframil: i want one user to be able to be able to upload into any direcotory in /home through ftp....and right now i have /home set to 777 so the user can do that...but i want to fix that and i dont know how
<11> signature16: say you have a couple non root users who you want to be file admins, make a group called fadmin , then add the users to that group
<14> which package contains nslookup?
<6> Good question! equery b doesn't tell us!!!
<5> signature16: yes, for example: chown root:fadmin /home && chmod 775 /home
<6> net-dns/bind-tools
<5> signature16: then gp***wd -a someuser fadmin
<5> signature16: setting rwx for the whole world isn't smart
<6> While with rpm you could do a simple: find / -name \*.rpm -print | while read package ; do rpm -q -l $package | grep -q -s nslookup && print $package ; done
<11> signature16: then : cd /home && for i in *;do chown -Rv $i:fadmin $i && chmod -Rv ug+rwX $i;done
<5> signature16: bind-tools
<5> oh doh, it was answered
<9> kerframil: what does chown root:fadmin /home do?
<15> hey, I try to emerge --update -p world and I get this
<15> !!! ARCH is not set... Are you missing the /etc/make.profile symlink?
<15> !!! Is the symlink correct? Is your portage tree complete?
<14> pjb kerframil: thanks
<14> cheers
<11> Visseroth: emerge --sync ??
<5> signature16: sets the owner to root (as it already should be) and the ***ociated group to fadmin (create/call the group whatever the heck you like)
<6> signature16: It changes the owner and the group of /home.
<15> well crap
<13> Visseroth: are you missing the /etc/make.profile symlink?
<13> Visseroth: Is the symlink correct?
<15> if that is it
<13> Visseroth: Is your portage tree complete?
<15> I dunno
<5> Visseroth: run "readlink /etc/make.profile"
<15> ager ~ # readlink /etc/make.profile
<15> ../usr/portage/profiles/default-linux/x86/2006.0
<13> Visseroth: that looks good
<15> ok, running emerge --sync right now
<13> Visseroth: have you done an emerge --sync lately?
<5> Visseroth: then your portage tree may be damaged/incomplete/non-existent. do you have the directory in question (minus the "../" part)?
<5> yeah, syncing is a good plan :)
<15> dunno yet, let me finish emerge --sync. but ya know, thanks a ton guys
<16> if i were to create a seperate partition for /usr/portage and /var how big should i make them? i use ccache with 4G as the size limit
<12> So I have 10's of 1000's of files on a Raid5 Array, updatedb takes hours to catalog everything on it. Any good alternative cataloging utils?
<13> seren: how do you plan to put two directories on a single partition?
<13> seren: or do you mean two partitions?
<16> seperate partitions
<13> seren: ah, ok
<16> 'and' ;)
<13> 'a separate partition'
<13> :-)
<17> hi all, is there any doc available on upgrading to gcc 4.1
<16> 6gb for /var and 1G for /usr/portage?
<17> ?
<6> My recently installed gentoo has 1.3G in /usr/portage and 188M /var
<16> pjb, i use ccache, wouldnt i need more space?
<13> seren: just keep in mind that /var/tmp/portage/ is used when compiling
<6> I don't know ccache. But anyways, more space will be needed.
<18> when you emerge any given package, there are often warnings on having to run or configure something which is nice. My question is if you run emerge -uDN world, how does one know what needs to be configured or ran?
<16> bilboed, first off you need glibc 2.4
<6> In particular, in /var you can fill 4GB easily, with spool and log
<13> so a large program might use up a good bit of space (e.g. OOo)
<8> Question: I'm currently installing, should I select Stage 3 since I have an internet connection, or use the last step, 3+GRP?
<6> biohazrd: there's etc-update
<11> biohazrd: redirect all output to a file, parse it afterwards
<6> Personnally, I just find /etc -name ._cfg0\*
<18> yes i know about et-update, however warnings for example for running eslect opengl after emerging nvidia
<17> seren, which is masked I suppose
<18> or gcc-config and source /etc/profile for gcc
<16> bilboed, its not in poetage
<6> Yes, for this you need to parse the output.
<11> pjb: he is talking about the ebuild warnings and post install instructions. not cfg updating....
<16> !google gentoo forums glibc 2.4
<8> any recommendation?
<19> http://forums.gentoo.org/viewtopic-t-309752.html
<6> SuSE sends emails to root!
<17> seren, thx
<8> would using stage 3 from the internet be a better choice then choosing stage 3+GRP?
<11> pjb: so does gentoo
<16> bilboed, once you get glibc 2.4 you cant go back
<9> does gp***wd -a user group add a user to a group?
<6> brianw: really? Then we only need to read our messages :-) Othewise, If the warning and post install instructions are issued with a special command (ewarn?), then it could be easy enough to redirect them to mail.
<17> seren, good to know. Once glibc 2.4 is emerged, the other software will still work with the new glibc ?
<16> i was planing on putting /var on reiser4 ;/ and /usr/portage on reiserfs
<11> pjb: the ~arch portage logs this info, imo it should of always been logged...
<20> I know it's off topic: does anyone know of a cell phone chat room? (on freenode or elsewhere?)
<21> signature16: yes, check man gp***wd for more information
<16> bilboed, so far i havent heard of many problems and i have been using it for a bit now 2.3.9X that is
Return to
#gentoo
or
Go to some related
logs:
centos libipt_comment.so
int to string in php
#web
#linux
UnboundLocalError: local variable 'gid' referenced before assignment
compiz-experimental
#suse
NatureTM irc
howto wipe mbr dapper
bash_history histsize gentoo
|
|