logo
IRC Archive / Freenode / #debian / 2010 / January / 25 / 4
echa
the same package
there is only one
jelly-home
echa: pastebin the actual error
echa
the file exists in only one package
it's hard to paste as it's on another system with no internet
i will copy to usb hang on
paissad
squisher, yes, if i remove the background option of the program , everything's ok !
squisher, you're a great guy :)
squisher
paissad, you're welcome
jelly-home
echa: if it's the exact package, things should work. If it's some other package, read the policy: http://www.debian.org/doc/debian-policy/ch-relationships.html#s-replaces
Devastator
jelly-home I see that you didn't sleep yet, can I ask you a simple question about raid?
jelly-home
maybe :-)
Devastator
jelly-home is it hard to setup raid 10 on linux?
jelly-home
not any harder than elsewhere
echa
http://pastebin.com/d1fd38253
jelly-home: i'm not sure what you mean
jelly-home: the file exists in only one package
jelly-home
pkg: error processing /var/cache/apt/archives/pack-eclipse_0.01_all.deb (--unpack): trying to overwrite `/opt/eclipse/about.html', which is also in package nirs-eclipse
paissad
btw, i would like to know how to manage the order of start of init.d scripts ... for example how to ensure that script_a will be run before script_b ?
jelly-home
echa: you have "pack-eclipse" and "nirs-eclipse"
echa
this is a mistake
abrotman
you made it!
jelly-home
echa: you'll have to either purge the old package first, or add Conflicts: and Replaces: nirs-eclipse if you want it to be automatically removed, or just Replace: if you want that file overwritten but keep other bits of the old package.
echa
but it isn't installed
jelly-home
echa: yes it is.
echa
i purged everything from synaptic
and checked in /top
jelly-home
it's unpacked at least
echa
in /opt
ok
jelly-home
echa: dpkg -l nirs\*
echa
no packages found matching
jimcooncat
I'd like to know if there is an option where I could have "group homes", where a home directory is provided for each group formed by manual input. Anyone heard of a setup like that?
jelly-home
echa: dpkg doesn't make mistakes... users do!
echa
yes
but where
that's the hard part
jimcooncat
I suspect this would be monkeying with PAM?
Rienzilha
dunno
jelly-home
echa: nothing in dpkg -l nirs\*, or dpkg -l nirs-eclipse?
Rienzilha
add them to a group
make the directory owned by that group
and set the homedir for each user to that directory
echa
jelly-home, no
jelly-home
echa: ls -l /var/lib/dpkg/nirs-eclipse*
Devastator
jelly-home I asked the question because I'm thinking about building a server using raid 10, but never used it under linux.. some say raid 10 is the best
jimcooncat
I guess I could make my own front end to addgroup?
Mr_Queue
any aptitude guru's on hand? It has been forever since I've been outside of Debian. However, I just set up an Ubuntu host inside my Sid machine and noticed a pretty big different in the responsiviness when using aptitude in ubuntu. what gives?
echa
nothing in there
something is wrong with my packages
i think
the control files are wrong
lizk
in grub2 what is the number of first logical partition?
jelly-home
echa: that's a real possibility, too. The clocks on your machines don't match either.
echa
yes, i noticed
one is a VM, its clock is always wrong
jelly-home
eh, make it ntpdate once a minute in cron
Mr_Queue
is there some large cache aptitude/dpkg is reading over I don't know about?
jelly-home
... once you have network
echa
jelly-home, no internets
jelly-home, no internets ever...
jelly-home
Mr_Queue: I guess, if their aptitude is faster, you'll have to ask them?
echa: so make a local time server, or install guest additions, or just do date -s "correct time"
jimcooncat
Mr_Queue: maybe one of the repositories you use added a load of packages?
sparr
are there any open source multi-protocol IM clients that support video on one or more protocols?
jelly-home
echa: also, don't expect help from me if you paste munged data, it's too much hassle to fix the real issue without the user muddying it even more
abrotman
Mr_Queue: you sohuld really ask #ubuntu
jelly-home
"Dear #ubuntu, why is Ubuntu's aptitude so much faster than Debian's?"
Mr_Queue: possibly a better question for #debian-devel than here, they might know if ubuntu has patches applied
abrotman
or ask the ubuntu devs .
jimcooncat
how do I find what pam modules debian stable has available?
Mr_Queue
jelly-home: good point.. I had initally thought that perhaps it had something to do with the age of my Sid install..(years) over the fresh vhost.
err.. vclient
jelly-home
Mr_Queue: you might have lots of crap in the aged installation sources.list
Mr_Queue
abrotman: why? I don't 'run' ubuntu really.. it is only a vm built just the other day, and the slack seems to be debian.
jelly-home: well, rather 'clean' sources.. main contrib non-free and multimedia..
jelly-home
Mr_Queue: we don't know why aptitude is slow as hell, we just live with it
abrotman
Mr_Queue: it's their patches, not ours .. ask them how they altered theirs to be faster
jelly-home
and probably a -dev channel of some sort rather than just #u
Mr_Queue
abrotman: well that seems to answer the question. with it is know and accepted that our aptitude is slower..
*known
DogBoy
slower how?
abrotman
Mr_Queue: good, conversation over?
Mr_Queue
abrotman: yeah sure I suppose so....
abrotman
Mr_Queue: really, you're asking the wrong people
tking
What would debian do with 1000 USD if one were to give them that much?
jelly-home
tking: TIAS
abrotman
tking: you'd donate the money to SPI
Mr_Queue
tking: pay for some bandwidth? add to the endowment?
tking
I wanted to setup a FSF scholarship for college students
but then my wife took a bunch of stuff
:|
Mr_Queue
heh- they're good for that eh
jelly-home
tking: you seem to be overestimating the value of US Dollar by a lot
TrueDD
Good night all
[bintut]
(Action) waves
tking
jelly, no I had a lot more than a grand. \
jelly-home, *
Mr_Queue
:)
[bintut]
i am having an OOM that kills my running Squid 2.6STABLE5 eventhough i have a 16GB RAM. any advice? how can i eliminate the OOM problem on my system? thanks.
OldParr
please how do i incorpore filesystem NTFS to gparted in debian it only shows ext2 ext3 and swap?
jelly-home
[bintut]: 32 or 64bit debian?
nolimitsoya
is there a live install cd with networkmanager enabled?
jelly-home
OldParr: install the Suggested: packages of gparted. Notably, ntfsprogs
OldParr
jelly-home: thanks
nike
jelly-home: What do you think of this: mdadm --assemble /dev/md0 /dev/sda1 /dev/sdb1 /dev/sdc1 /dev/sdd1 /dev/sde1 /dev/sdf1
jelly-home
nike: sounds worth a try... is sdb alive?
nike
jelly-home: best way to check?
lizk
in grub2 what is the number of first logical partition?
jelly-home
nike: hdparm -tT? smartmontools?
dvs
lizk: 5
bododo
hi guys
jimcooncat
is there anything special about a user or group name that starts with an underscore?
nolimitsoya
ok the, does the kde install come with networkmanager or will i need more cds?
*then
bododo
it seems that udev has problems in sid?
jelly-home
nike: if you take a look at superblock in sda1, you'll see there was no sdb1 in there the last time it was working, either
bododo
It shows me a whole lot of errors (warnings I think)
and mouseemu isn't working
dvs
bododo: about conf?
nike
jelly-home: Odd. Right. Okay, scary. If I erased sdf during that stunted 1-2% rebuild process when it was sda, sdb, etc....
bododo
when I login to X, I have no keyboard
dvs: ? :)
jelly-home
nike: you'll probably have some crap on sdf1
bododo
dvs: ah yes! the warnings were about conf I think
nike
jelly-home: But look at this: http://pastebin.ca/1764721
It shows hdparm output as well as mdadm scanning sdb1
dvs
bododo: they're just warnings.
bododo
is there a way I can tune/config udev or something?
jelly-home
nike: note the difference in dates on sdb1 and other sd?1
nike
ah
bododo
dvs: yes, but I think there's a pb with it
nike
jelly-home: Okay so it might have died on Nov 3?
That hdparm command didn't have much info it seems.
Should I do another hdparm command or install smartmontools?
OldParr
please how do i incorpore filesystem FAT to gparted in debian it only shows ntfs ext2 ext3 and swap?
jelly-home
nike: at this point I'd either contact a data recovery company and explain in detail what was done, or try assembling sda1, sdc1-sdf1 and see what the filesystem looks like
nike
jelly-home: Wow. Okay....
jelly-home: What do you think of this: mdadm --assemble /dev/md0 /dev/sda1 /dev/sdc1 /dev/sdd1 /dev/sde1 /dev/sdf1
jelly-home
OldParr: look at the status of gparted package. dpkg -s gparted. Then extrapolate from that.
nolimitsoya
nike, specify a metadata version, level and layout as appropriate
jelly-home
OldParr: the answer is similar to your previous question.
nolimitsoya
nike, oops, never mind :)
OldParr
jelly-home: thanks
jelly-home
nike: I think I'd have to rtfm to see if there are any --read-only options to be added
pm2
I'm having a strange problem with the way mail is being sent to certain users on my domain. I have a postfix/procmail/spamassassin setup. A user has an account which is set to deliver to a virtual mailbox, and to forward to a blackberry email address. When I send mail to that user, he received 9 copies of the message. Log files are here: http://pastebin.com/m4b7369e2
nike
jelly-home: yeah read-only would be amazing, that's really the reason I'd want to use force but
pm2
This started after I setup procmail and spamassassin to do spam filtering. So far I've only noticed it on this one user's account, I setup another account to forward to a non-blackberry address and also deliver to a virtual mailbox, and I don't have the problem. Any ideas?
OldParr
jelly-home: why do you use dpkg instead of aptitude?
jelly-home
OldParr: why not?
nike
jelly-home: I do not see one in the subsection of switches for assemble nor the general non-mode specific switches
OldParr
jelly-home: internet says that aptitude is the best package manager for debian
jelly-home
OldParr: I don't see the point. Is there a shorter way to do dpkg -s packagename with aptitude?
OldParr
jelly-home: i will investigate that question i really do not know
g00dtimes
pm2: for clarity, you send one email to w@example.com and it gets delivered 9 times to sw@vzw.blackberry.net?
jelly-home
neither do I, so I use what I know works
g00dtimes
s/w@example.com/sw@example.com
pm2
g00dtimes, let me double check...
jelly-home
nike: also. That hdparm test showed hdparm was unable to read 2MB off the disk... so sdb has issues.
Timing buffered disk reads: read(2097152) returned 1445888 bytes
nike
jelly-home: Probably enough to kill it's position in the array?
Either needs to be rebuilt, or... well does that mean it's definitely dead?
And by dead I mean: is that 2MB of space physically shot?
pm2
g00dtimes, I send one email to sw@example.com, and 9 copies end up at both sw@vzw.blackberry.net and the virtual maildir for sw@example.com
jelly-home
nike: it certainly seems so
g00dtimes
pm2: what are you using for your virtual maps?
nike
jelly-home: Okay well I am scared I may have overwritten sdf's data from before when I had no sda1, sdb1, etc.
best way to verify would be to mdadm --assemble I think
leaving out sdb I guess
pm2
g00dtimes, a pgsql backend.. let me post main.cf...
jelly-home
nike: the safest thing to do at this point might be: get a new disk to use instead of current sdf. Then assemble a degraded array with sd{a,c,d,e}, and let it rebuild on the replacement for sdf
nike
jelly-home: Probably will add --run actually, that attempts to start the arry even if less drives are specified than when it was last working. So if I only specify 5 drives and leave out sdb1...?
jelly-home
nike: this will work IFF no other disks have bad sectors/read errors
nike
jelly-home: No no, this is RAID-5 though.
If I lose 2 disks I thought I am dead in the water.
pm2
g00dtimes, main.cf at http://pastebin.com/m44814f3a
nike
If my sdb is physically dead... don't I need to try my damndest to recover sdf?
jelly-home
nike: the array seems to be using 5 devices.
nike
jelly-home: It's definitely a 6 drive RAID-5 array
jelly-home
oh, I misread
pm2
g00dtimes, I manually ran the sql query for sw@example.com - it returns two rows of sw@example.com and sw@vzw.blackberry.net
jelly-home
nike: then you're fu*ked already
nike
I've got sdb maybe dead, and sdf maybe overwritten
g00dtimes
pm2: also, hang a few "-v" off the end of the line in master.cf that contains the virtual definiton and then reload postfix
nike
jelly-home: What if it rebuilds with sdf?
mdadm --assemble /dev/md0 --run /dev/sda1 /dev/sdc1 /dev/sdd1 /dev/sde1 /dev/sdf1 ??
jelly-home
nike: s/rebuilds/assembles/? Some of the data might be crap.
don't insist on --run, first see if it assembles at all
nike
jelly-home: Well okay, I will. But now I'm looking at incremental mode.
g00dtimes
pm2: can you post pgsql-virtual-mailbox.cf as well?
jelly-home
nike: there's nothing incremental to be done here.
nike
Not sure if I want to just do that instead of a straight assemble
pm2
g00dtimes, certainly, one sec
nike
ah okay, I just though that the test incremental mode does might be nice, but okay.
jelly-home
nike: it's the same thing
nike
You think I should just run the assemble command without sdb?
jelly-home
all of your non-bad device are online and connected
g00dtimes
pm2: first off, crank up the verbosity on virtual. next, resend a message. it looks like the lookup is returning >1 result, thus delivering it more than once.
jelly-home
nike: definitely, sdb is a) dead, and b) data on it is 2 months old
nike
It'll know it's RAID-5 and try and run the array without sdb etc.