1331090870 J * clopez ~clopez@82.25.60.213.dynamic.mundo-r.com
1331094271 Q * clopez Ping timeout: 480 seconds
1331095166 J * notau ~notau@14-201-81-158.static.tpgi.com.au
1331099606 N * Bertl_zZ Bertl
1331099616 M * Bertl morning folks!
1331101866 J * ghislain ~AQUEOS@adsl2.aqueos.com
1331104552 M * hijacker morning
1331108463 J * petzsch ~markus@dslb-092-078-115-163.pools.arcor-ip.net
1331109628 Q * petzsch Quit: Leaving.
1331111789 J * petzsch ~markus@dslb-092-078-115-163.pools.arcor-ip.net
1331112843 M * ex hello, I see that in kernels >= 3.0.20-vs2.3.2.2-beng  memory stats in /proc/virtual/X/limits are not working
1331112857 M * ex VM/RSS counters are 0
1331112879 M * ex is it bug or I should use other source (cgroups etc)
1331113120 M * Bertl is expected behaviour, Linux-VServer switched to cgroups
1331113155 M * Bertl (but that was around 2.6.30 actually :)
1331113591 M * ex it was working with 2.6.32, ok, anyway, I will use cgroups memory.stat
1331113630 M * ex there is no VM equivalent there right?
1331113649 M * Bertl VM?
1331113678 M * Bertl ah, well, there is memory + swap now, but no VMA sum
1331113768 M * ex vserver-stat show VSZ value, but i cannot figure out how to calculate it from cgroups
1331113800 M * Bertl it's no calculated from there, it's just the sum over all processes
1331113942 M * ex ok, got it, thanks for explanation!
1331113979 M * Bertl np
1331114044 M * daniel_hozac VSZ is never an interesting value though
1331114942 M * Bertl daniel_hozac: was there a change in the documentation or the way it is built between pre3025 and pre3029?
1331114987 M * Bertl I'm asking because the former builds fine on the mageia build system but the latter bails out when generating the docu (both build fine here, so I'm a little confused atm)
1331116424 M * Mr_Smoke 'morning
1331116773 M * Mr_Smoke I have a question about moving vservers across physical hosts
1331116803 M * Mr_Smoke The hosts have the same setup and the Vservers live on LVM volumes
1331116826 M * Mr_Smoke when VS1 is running on HOST1, the volume is rsynced on a regular basis across to HOST2
1331116831 M * Mr_Smoke Now here comes the question
1331116875 M * Mr_Smoke Is there a way to speed up the rsyncs in the hot-rsync/shutdown on 1/cold-rsync/start on 2 process?
1331116900 M * Mr_Smoke the guest containing maildirs takes ages to rsync because of the amount of files I think
1331116972 M * daniel_hozac Bertl: hmm, no.
1331116983 M * Bertl depending on the kernel and disk I/O subsystem, you can speed up maildir like storages significantly by selecting the proper filesystem and I/O tuning
1331116984 M * daniel_hozac Bertl: do you have a log?
1331117106 M * Bertl give me a few minutes, I simply resubmit it, maybe it works now, if not, it will produce a recent log :)
1331117255 M * Mr_Smoke Bertl: just to clarify, the maildir performance as such is ok, it's just the transfer I need sped up. I guess your suggestion still stands though
1331117295 M * Bertl yes, usually most of the 'delay' with many small files is caused by metadata updates
1331117315 M * Mr_Smoke Are you thinking XFS ?
1331117323 Q * Romster Quit: Geeks shall inherit properties and methods of object earth.
1331117335 M * Bertl jfs/xfs is a good approach from the filesystem PoV
1331117341 M * Mr_Smoke Got it
1331117347 M * Mr_Smoke I guess I'll have to try
1331117361 M * Bertl but the underlying I/O subsystem can usually be tuned as well
1331117370 M * Bertl one major problem is read-ahead
1331117411 M * Bertl if your meta data is scattered all over the disks, you definitely do not want to read ahead 128k for each 4k (or even 512byte) metadata block
1331117444 M * Bertl daniel_hozac: http://pkgsubmit.mageia.org/uploads/failure/cauldron/core/release/20120307104620.bertl.valstar.27206/
1331117449 M * Mr_Smoke Hm. And that (scattered metadata) depends on the FS I take it?
1331117465 M * Bertl yes, filesystem and filesystem configuration
1331117492 M * Bertl daniel_hozac: specifically: http://pkgsubmit.mageia.org/uploads/failure/cauldron/core/release/20120307104620.bertl.valstar.27206/log/util-vserver-0.30.216-0.pre3029.2.mga2/build.0.20120307104622.log
1331117542 M * Mr_Smoke 'k
1331117552 M * Bertl daniel_hozac: maybe it hits a cpu limit or similar, the funny part is it seems to build fine on x86_64 (via buildsystem) as well
1331117553 Q * Mr_Smoke Read error: Connection reset by peer
1331117575 J * Mr_Smoke ~smokey@layla.lecoyote.org
1331117607 M * daniel_hozac Bertl: weird.
1331117615 M * daniel_hozac i see no error except the make error either.
1331117863 M * Bertl okay, thanks for the information!
1331117895 M * Bertl off for a nap now ... bbl
1331117907 N * Bertl Bertl_zZ
1331120723 J * xe ex@valis.net.pl
1331120735 Q * ex Read error: Connection reset by peer
1331120735 N * xe ex
1331121189 J * clopez ~clopez@82.25.60.213.dynamic.mundo-r.com
1331122107 J * ensc|w ~ensc@www.sigma-chemnitz.de
1331122257 Q * ex Ping timeout: 480 seconds
1331122410 M * Mr_Smoke hi again
1331122418 M * Mr_Smoke I'm about to try stuff with /etc/vservers/vserver-name/scripts 
1331122453 M * Mr_Smoke question: what variables/arguments are passed on to the scripts ? stuff like "what vserver am I being called for" is what I'm after
1331122463 M * Mr_Smoke so that I could write a generic script that does iptables traffic accounting
1331122563 M * Mr_Smoke Hm apparently vserver name = $2
1331122568 M * Mr_Smoke worth a try :)
1331123100 J * ex ex@valis.net.pl
1331124080 Q * petzsch Quit: Leaving.
1331124444 Q * ex Ping timeout: 480 seconds
1331124737 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com
1331124965 J * ex ex@valis.net.pl
1331127526 Q * notau 
1331127920 Q * Aiken Remote host closed the connection
1331128159 Q * clopez Ping timeout: 480 seconds
1331128677 J * notau ~notau@14-201-81-158.static.tpgi.com.au
1331129584 N * Bertl_zZ Bertl
1331129593 M * Bertl back now ...
1331129873 M * ghislain yes mr_smoke from mémory $1= scr'ipt name $2= guest name
1331130307 M * Mr_Smoke ghislain: yep, figured it out, thanks :)
1331130324 M * daniel_hozac also says in doc/configuration.xml
1331131001 M * ghislain would it be possible to add it to the great flower page ?
1331131026 M * ghislain with the "base" parametrers for cgroups default :) ?
1331131529 M * ghislain daniel would it be possible to link sub part of the great flower page to wiki pages ?
1331131554 M * ghislain i mean with some link tied to the wiki people would e able to link configuration exemple and knowledge
1331131571 M * Bertl the flower page is built from the sources
1331131589 M * Bertl I don't think it is a good idea to put wiki crosslinks into source code
1331131631 M * ghislain like /etc/vservers/.defaults/cgroup  would link to  http://linux-vserver.org/util-vserver:Documentation:-etc-vservers-.defaults-cgroup
1331131648 M * ghislain ah
1331131711 M * ghislain but the base cgroup  parameters is in the source but not in the flower :(
1331131728 M * ghislain i think that would be a good idea to have such way to enhance the docs
1331131750 M * ghislain duplicating the flowerpage in the wiki would be not good because it changes
1331131767 M * ghislain so make it link wiki page sounded like a good idea to me ^^
1331131850 M * ghislain but all my ideas are not considered good so there is a slight chance that it is not :p
1331131914 J * derjohn_mob ~aj@87.253.171.214
1331131991 M * Bertl ultimately it's up to daniel_hozac :)
1331132132 M * ghislain eheh :) the goal would be to ease people path with exemple , i think a template for exemples with version of the utils and the kernel is also needed if not it will become just like some wiki parts that are obsolete and confuse people instead of helping them
1331132151 M * ghislain i'd like to put some time on  it if i find some i will !
1331132874 J * petzsch ~markus@dslb-092-078-115-163.pools.arcor-ip.net
1331134315 N * ensc Guest5381
1331134325 J * ensc ~irc-ensc@p54ADED04.dip.t-dialin.net
1331134734 Q * Guest5381 Ping timeout: 480 seconds
1331134889 J * clopez ~clopez@155.99.117.91.static.mundo-r.com
1331135329 Q * grobie Ping timeout: 480 seconds
1331135450 J * grobie ~grobie@78.47.105.20
1331136243 J * dowdle ~dowdle@scott.coe.montana.edu
1331136384 Q * ncopa Quit: Leaving
1331136539 Q * grobie Ping timeout: 480 seconds
1331139144 M * Bertl off for now ... bbl
1331139155 N * Bertl Bertl_oO
1331139619 Q * notau Quit: Computer has gone to sleep.
1331139913 Q * Mr_Smoke Read error: Connection reset by peer
1331140054 J * Mr_Smoke ~smokey@layla.lecoyote.org
1331145648 Q * derjohn_mob Ping timeout: 480 seconds
1331145793 N * Bertl_oO Bertl
1331145799 M * Bertl back now ...
1331145840 Q * petzsch Quit: Leaving.
1331146822 M * clopez i have a question related to the warning "vxW: [»wall«,15601:#40001|40001|40001] did lookup hidden devpts:ffff880439255580[#40015,3] »/dev/pts/0« "
1331146830 M * clopez what does this messages means?
1331147554 M * Bertl it means that a process (wall, pid 15601) from context 40001 tried to access /dev/pts/0 which belongs to context 40015
1331147588 M * Bertl and this was successfully prevented by the kernel
1331148185 M * clopez this happens many times when I do a "vserver xxx enter" from the master
1331148206 J * derjohn_mob ~aj@88.128.42.203
1331148236 M * Bertl yep
1331148252 M * Bertl because you bring your ptys into the guest
1331148310 M * clopez i see
1331148318 M * clopez thanks for the clarification :)
1331148325 M * Bertl you're welcome!
1331149559 Q * clopez Ping timeout: 480 seconds
1331150578 J * hijacker_ ~hijacker@cable-84-43-136-96.mnet.bg
1331151263 J * petzsch ~markus@dslb-092-078-115-163.pools.arcor-ip.net
1331151268 Q * petzsch 
1331153635 J * Aiken ~Aiken@2001:44b8:2168:1000:21f:d0ff:fed6:d63f
1331154886 J * clopez ~clopez@82.25.60.213.dynamic.mundo-r.com
1331155295 Q * derjohn_mob Ping timeout: 480 seconds
1331155439 Q * clopez Ping timeout: 480 seconds
1331156618 Q * hijacker_ Quit: Leaving
1331158146 J * derjohn_mob ~aj@tmo-094-220.customers.d1-online.com
1331158795 Q * ghislain Quit: Leaving.
1331158971 J * notau ~notau@14-201-81-158.static.tpgi.com.au
1331161767 J * clopez ~clopez@82.25.60.213.dynamic.mundo-r.com
1331164467 J * bzed_ ~bzed@devel.recluse.de
1331164774 Q * bzed Ping timeout: 480 seconds
1331164777 N * bzed_ bzed