1326240541 Q * bergerx Quit: Leaving
1326240964 Q * geos_one Ping timeout: 480 seconds
1326241811 J * daniel_hozac ~daniel@c-bb3271d5.08-230-73746f22.cust.bredbandsbolaget.se
1326243878 J * geos_one ~chatzilla@chello080109195117.4.graz.surfer.at
1326243984 Q * dowdle 
1326247246 Q * fisted synthon.oftc.net larich.oftc.net
1326247246 Q * cuba33ci synthon.oftc.net larich.oftc.net
1326247246 Q * tolkor synthon.oftc.net larich.oftc.net
1326247246 Q * PowerKe synthon.oftc.net larich.oftc.net
1326247246 Q * transacid synthon.oftc.net larich.oftc.net
1326247246 Q * misc-- synthon.oftc.net larich.oftc.net
1326247270 J * fisted ~fisted@xdsl-87-78-216-107.netcologne.de
1326247270 J * cuba33ci ~cuba33ci@114-36-234-3.dynamic.hinet.net
1326247270 J * tolkor ~rj@tdream.lly.earlham.edu
1326247270 J * PowerKe ~tom@94-226-105-186.access.telenet.be
1326247270 J * transacid ~transacid@transacid.de
1326247270 J * misc-- ~misc@202.171.160.4
1326248224 J * fisted_ ~fisted@xdsl-87-78-218-123.netcologne.de
1326248431 Q * fisted Ping timeout: 480 seconds
1326249006 Q * nicholi synthon.oftc.net charm.oftc.net
1326249006 Q * hparker synthon.oftc.net charm.oftc.net
1326249006 Q * Romster synthon.oftc.net charm.oftc.net
1326249006 Q * thal synthon.oftc.net charm.oftc.net
1326249006 Q * jrklein synthon.oftc.net charm.oftc.net
1326249006 Q * Alex[fob] synthon.oftc.net charm.oftc.net
1326249020 J * nicholi ~nicholi@rrcs-76-79-196-34.west.biz.rr.com
1326249020 J * hparker ~hparker@2001:470:1f0f:32c:beae:c5ff:fe01:b647
1326249020 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326249020 J * thal ~thalunil@walledcity.de
1326249020 J * jrklein ~osx@2001:470:1f0f:572::250:160
1326249020 J * Alex[fob] ~alex@v1.fob.spline.inf.fu-berlin.de
1326251019 J * ghislain ~AQUEOS@adsl2.aqueos.com
1326251619 Q * ghislain Quit: Leaving.
1326256124 J * mike3 ~mike@S01060023bee96928.vs.shawcable.net
1326256166 Q * mike3 Excess Flood
1326256181 J * mike3 ~mike@S01060023bee96928.vs.shawcable.net
1326256502 Q * mike3 autokilled: Do not spam. Mail support@oftc.net with questions (2012-01-11 04:35:02)
1326257025 Q * AndrewLee Read error: Operation timed out
1326263282 J * sannes ~ace@cm-84.209.106.118.getinternet.no
1326264913 J * AndrewLee ~andrew@n201.enc.hlc.edu.tw
1326267043 J * Marbug ~Marbug@83.101.67.3
1326269943 J * ghislain ~AQUEOS@adsl2.aqueos.com
1326272169 Q * ensc|w Remote host closed the connection
1326272177 J * ensc|w ~ensc@www.sigma-chemnitz.de
1326272413 Q * Marbug Ping timeout: 480 seconds
1326273602 M * Guy- in retrospect, this was probably a side effect of me using mount --make-shared
1326274222 M * renihs Guy-, hmm how/what do you mean with namespace cleanup?
1326274275 M * Guy- renihs: when you start a guest, it receives a copy of the host's mount namespace; but that namespace is cleaned up by removing from it all filesystems whose final mountpoint lies outside the root of the guest
1326274304 M * Guy- renihs: it's possible to customise this somewhat in /etc/vservers (it can be turned off entirely and specific mountpoints can be excluded from it)
1326274322 M * renihs hmm trying to reproduce that all zfs gets umounted 
1326274371 M * renihs which i fail at or i did misunderstand
1326274542 M * renihs Guy-, not sure, can you provide steps howto reproduce that? for me, if i vnamespace mount a zfs dataset inside a guest it gives me expected behaviour
1326274685 M * Guy- renihs: zfsonlinux or zfs-fuse?
1326274697 M * renihs zfsonlinux ofc :)
1326274706 M * Guy- renihs: if zfsonlinux: create some zfs datasets in a hierarchy under, say, /foo
1326274717 M * renihs btw, friend just finished patching illumos grub to be able to boot from zfs if you need that
1326274728 M * Guy- e.g. zfs create -p -o mountpoint=/foo tank/foo/bar/baz
1326274737 M * Guy- no
1326274743 M * Guy- e.g. zfs create -p -o mountpoint=/foo tank/foo
1326274753 M * Guy- zfs create -p tank/foo/bar/baz
1326274766 M * Guy- and, say, zfs create -p tank/foo/x/y
1326274789 M * Guy- now, mount --make-shared /foo (I think this is where I screwed up)
1326274796 M * renihs oh hmm
1326274812 M * renihs i vnamespaced the zfs mount itself (to be mounted inside the guest)
1326274814 M * renihs will try that
1326274815 M * Guy- and create a guest that has /foo/bar/baz and /foo/x/y bind mounted somewhere
1326274825 M * renihs didnt use bind mounts
1326274831 M * renihs not sure if i want that?
1326274834 M * Guy- start that guest with namespace cleanup enabled (which is the default)
1326274837 M * renihs k
1326274853 M * Guy- and then check whether on the host /foo, /foo/bar, and /foo/x are still mountpoints
1326274908 M * Guy- (I have to use bind mounts because I'm sharing some directories among guests - the other option would be nfs)
1326275255 M * renihs hmm did that but hmm nothing happened what wasnt expected behaviour
1326275289 M * renihs added those mounts to the /etc/vserver/<guest>/fstab and rebooted, but they get mounted as expected
1326275325 M * renihs ah oh wait
1326275364 M * renihs forgot to mount --make-shared 
1326275378 M * renihs i get a completely different error though, the vserver refuses to start now
1326275424 M * renihs weird...
1326275459 M * renihs y, everything gets umounted :)
1326275476 M * renihs evil
1326275505 M * renihs so its definitve the mount --make-shared action that results in that
1326275515 M * renihs (which i forgot initially)
1326275649 Q * Guy- Ping timeout: 480 seconds
1326276235 M * renihs also evil that i cant umount the filesystem from the original vserver anymore :)
1326276300 J * Guy- ~korn@elan.rulez.org
1326277029 M * renihs guy, dunno what you red/missed from my wall of text :p
1326277050 M * renihs http://pastebin.com/3FqYZUE4
1326277062 M * renihs but y, mount --make-shared breaks it
1326277081 M * renihs + i cant umount the dataset with the vserver i tried it on anymore :)
1326277092 M * renihs guess need a reboot
1326278154 J * petzsch ~markus@dslb-094-222-072-215.pools.arcor-ip.net
1326281768 J * gucki ~gucki@80-218-125-247.dclient.hispeed.ch
1326283548 Q * petzsch Quit: Leaving.
1326284435 Q * Romster Quit: Geeks shall inherit properties and methods of object earth.
1326284652 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326284679 Q * Aiken Remote host closed the connection
1326285181 Q * Romster Quit: Geeks shall inherit properties and methods of object earth.
1326285195 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326285696 M * Guy- renihs: I didn't see your wall of text at all; pastebin says unknown paste ID
1326285712 M * Guy- renihs: did you also try without --make-shared?
1326285761 M * renihs Guy-, http://pastebin.com/JS8x34D5
1326285774 M * renihs Guy-, well nothing bad happened
1326285779 M * renihs until i added --make-shared :)
1326285784 M * renihs then everything broke
1326285811 M * renihs cant even umount the dataset from the vserver i tried it with :)
1326285815 M * renihs guess needs reboot
1326286002 M * Guy- renihs: OK, thanks for confirming :)
1326286019 M * renihs Guy-, yw :)
1326286081 Q * Romster Read error: Operation timed out
1326286083 M * renihs i am working hard to get Bertl interested in zfs btw :)
1326286093 M * renihs so far he resisted all my attempts to make him more curious :)
1326286127 M * renihs well working hard is an exaggeration i guess, i am still diggin into it myself, but dedup seems very nice for vservers, i prefer it above unification tbh
1326286144 M * Guy- renihs: I was thinking about that, and I'm not sure it's better
1326286167 M * Guy- renihs: if you consider a shared library, of which many copies exist
1326286190 M * Guy- renihs: if those have the same inode number, the kernel will map a single copy into RAM
1326286215 M * Guy- renihs: I doubt this is the case if the inode numbers are different, but the disk blocks are the same
1326286233 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326286257 M * Guy- renihs: also, there is the substantial overhead of deduplication - I was also very excited about it at first, but now I find myself disabling it almost everywhere and staying with compression (which is practically free)
1326286285 M * renihs hmm well, its costs *alot* memory and also isnt too light on cpu
1326286294 Q * Romster 
1326286299 M * renihs however, if you for example have ssds (limited storage) then it gets interesting again
1326286309 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326286310 M * Guy- it's the memory usage that makes it prohibitively expensive for me
1326286319 M * Guy- I have SSDs, which I use as ZIL and L2ARC
1326286330 M * renihs and unification i guess is better for libs/bins but everything else dedup wins, its situational i guess
1326286353 M * Guy- I'm still trying to figure out how to determine how much RAM is needed to track L2ARC contents
1326286363 M * renihs well i am currently testing it on 12 vservers, and i get a 8:1 dedup ratio
1326286374 M * Guy- yes, dedup is better for /var style stuff
1326286375 M * renihs hmm there was a nice script to make it more human readable hmm 
1326286392 M * Guy- I found arcstats.pl and some similar scripts, but all for Solaris and FreeBSD
1326286411 M * Guy- my perl-fu isn't up to porting them to the /proc interface of zfsonlinux
1326286434 M * renihs http://groups.google.com/a/zfsonlinux.org/group/zfs-discuss/browse_thread/thread/2a822bb27a1cdce5/e612e9772a64d34f?lnk=gst&q=arcstat.pl#e612e9772a64d34f
1326286438 M * renihs arcstat for linux :)
1326286446 M * renihs you need also kstat but else its nice
1326286485 J * BenG ~bengreen@cpc10-aztw24-2-0-cust114.aztw.cable.virginmedia.com
1326286497 M * Guy- thanks, this will come in handy
1326286500 M * renihs https://github.com/zfsonlinux/linux-kstat
1326286505 M * Guy- yes, I see it
1326286583 M * renihs and https://github.com/zfsonlinux/arcstat 
1326286586 M * renihs obviously :)
1326286599 M * renihs i think i linked the wrong discussion
1326286633 M * renihs but y, i once had a nice paper on determing the RAM stuff for dedup etc
1326286645 M * renihs cant find it now, will search later more 
1326287289 Q * Romster Read error: Connection timed out
1326287324 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326288706 Q * BenG Quit: I Leave
1326290638 Q * FireEgl Quit: Leaving...
1326290717 J * petzsch ~markus@dslb-094-222-072-215.pools.arcor-ip.net
1326293031 J * fisted ~fisted@xdsl-87-78-211-96.netcologne.de
1326293206 Q * fisted_ Ping timeout: 480 seconds
1326293620 Q * Romster Read error: Connection timed out
1326293647 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326294670 Q * Romster Read error: Connection timed out
1326294703 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326295077 J * dowdle ~dowdle@scott.coe.montana.edu
1326295252 N * ensc Guest23707
1326295262 J * ensc ~irc-ensc@p54ADEB32.dip.t-dialin.net
1326295520 Q * kshannon Remote host closed the connection
1326295673 Q * Guest23707 Ping timeout: 480 seconds
1326295846 J * kshannon ~kris@122.252.14.166
1326295877 Q * gucki Remote host closed the connection
1326296215 N * Bertl_zZ Bertl
1326296220 M * Bertl morning folks!
1326296258 M * dkg morning, Bertl 
1326296428 Q * Romster Read error: Operation timed out
1326296450 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326296686 Q * ghislain Quit: Leaving.
1326296693 J * ghislain ~AQUEOS@adsl2.aqueos.com
1326297116 M * Bertl Guy-, renihs: I'd start testing zfs on an unpatched (not Linux-VServer patched :) kernel with different namespaces (i.e. unshared mount and filesystem spaces) ... because I presume it will fail horribly in this setup
1326297219 M * Bertl regarding dedup: I doubt that you get an actual benefit over unification, on the contrary, I presume it will use up more memory in almost any case
1326297359 M * renihs Bertl, hmm good idea, will see if can play around with namespaces abit tomorow
1326297373 M * renihs Bertl, regardind memory usage: certainly :p
1326297412 Q * Romster Read error: Connection timed out
1326297434 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326297794 Q * Romster Read error: Operation timed out
1326297816 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326297989 M * Bertl renihs: so where is the point in using it then? do you have a setup where memory is cheap/available but disk space a scarce resource (and still zfs the proper filesystem)?
1326298030 M * renihs Bertl, y 
1326298053 M * renihs also i like zfs data integrity features :)
1326298091 M * Bertl I agree, the data and metadata checksums are nice, care to elaborate regarding dedup?
1326298119 M * renihs but i have one deployment which consists mainly of ssds and space is rather limited, still havent decided about unification/dedup, also i will be testing zfs on linux for at least half more a year before deciding
1326298191 M * renihs Bertl, well, actually, due to user-mammal behaviour quiet alot files (non bins/libs etc) end up being duplicates 
1326298198 M * renihs on a specific setup i have 
1326298227 M * renihs with only bits and pieces of those differing at the most 
1326298244 M * renihs and for complicated reasons i cannot remove that or replace that with something more "sharelike"
1326298292 M * renihs besides that, i like zfs in general tbh
1326298299 M * renihs especially snapshot features etc :)
1326298308 M * renihs in one tool i mean 
1326298411 M * renihs but y, in general i admit that for vservers itself, zfs dedup features offer no real reason above unification 
1326299445 M * Bertl okay, thanks!
1326299450 M * Bertl off for now ... bbl
1326299455 N * Bertl Bertl_oO
1326301678 J * bonbons ~bonbons@2001:960:7ab:0:8887:a0bf:7344:5139
1326301837 Q * Romster Read error: Connection timed out
1326301860 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326302891 Q * Romster Read error: Connection timed out
1326302925 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326303905 Q * Romster Read error: Connection timed out
1326303929 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1326306900 Q * fisted Quit: brb
1326307212 J * fisted ~fisted@xdsl-87-78-211-96.netcologne.de
1326307505 J * Marbug ~Marbug@83.101.67.3
1326309527 Q * fback Ping timeout: 480 seconds
1326309742 J * fback fback@red.fback.net
1326310744 J * ryker ~jalberts@c-67-176-243-86.hsd1.in.comcast.net
1326312164 P * petzsch 
1326312947 Q * ghislain Quit: Leaving.
1326314260 J * Aiken ~Aiken@2001:44b8:2168:1000:21f:d0ff:fed6:d63f
1326317227 Q * cuba33ci Read error: Connection reset by peer
1326317326 J * cuba33ci ~cuba33ci@114-36-230-145.dynamic.hinet.net
1326318403 Q * bonbons Quit: Leaving
1326322199 Q * ryker Quit: ryker
1326323964 J * ghislain ~AQUEOS@adsl2.aqueos.com
1326324034 Q * Marbug Ping timeout: 480 seconds
1326326351 M * ghislain http://lwn.net/Articles/465074/ i don't know for you but this remind me something ! lol