1396569601 Q * fisted Remote host closed the connection
1396569612 J * fisted ~fisted@xdsl-87-78-182-54.netcologne.de
1396571091 J * undefined ~undefined@00011a48.user.oftc.net
1396571271 M * undefined patch-3.10.35-36.xz broke patch-3.10.33-vs2.3.6.8.diff (specifically this commit: http://git.kernel.org/cgit/linux/kernel/git/stable/linux-stable.git/commit/?id=0a0ae7b3fb0fb301da83f7c7da38807c76b2b869)
1396571291 M * undefined i'm working on fixing the vserver patch up to accommodate the changes
1396571322 M * undefined and i have a question regarding inode->i_vflags
1396571380 M * undefined it appears that i_vflags can contain V_BARRIER, V_COW, V_VALID, and V_XATTR
1396571427 M * undefined but searching through the linux code i never see V_VALID or V_XATTR set (or at least not referenced except defined in include/linux/fs.h)
1396571703 M * undefined set me study the new set_mask_bits to see if V_VALID and V_XATTR even matter in ext4_set_inode_flags (as they don't currently)
1396571911 M * undefined nevermind, set_mask_bits maintains the value of all bits not set by mask, so V_VALID and V_XATTR are not set/reset
1396572381 J * thierryp ~thierry@home.parmentelat.net
1396572863 Q * thierryp Ping timeout: 480 seconds
1396575963 Q * zerick Remote host closed the connection
1396575982 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:fddf:9561:d3cb:7abc
1396576465 Q * thierryp Ping timeout: 480 seconds
1396579582 J * thierryp ~thierry@home.parmentelat.net
1396580066 Q * thierryp Ping timeout: 480 seconds
1396583181 J * thierryp ~thierry@home.parmentelat.net
1396583663 Q * thierryp Ping timeout: 480 seconds
1396586782 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:d1a:25b7:e2c7:1907
1396587265 Q * thierryp Ping timeout: 480 seconds
1396587619 Q * undefined Quit: Closing object
1396587918 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:e120:112e:d65c:180e
1396589405 Q * thierryp Remote host closed the connection
1396590850 Q * wmp Read error: Operation timed out
1396591438 M * Bertl morning folks!
1396591544 J * wmp ~wmp@auburn.sored.pl
1396593725 J * noob123 ~kvirc@cpe-31-15-133-178.cable.telemach.net
1396593763 J * Ghislain ~aqueos@adsl1.aqueos.com
1396593810 N * noob123 snixor
1396594478 M * fback morning Bertl
1396594847 J * thierryp ~thierry@zebra.inria.fr
1396595046 M * hijacker morning
1396595295 M * Bertl off for now ... bbl
1396595314 N * Bertl Bertl_oO
1396596438 Q * Aiken Ping timeout: 480 seconds
1396598959 J * jakobsg ~chatzilla@62.116.207.125
1396598998 M * jakobsg Is bflag NET_RAW what i need to do tcpdump on a vserver guest?
1396599080 M * jakobsg if yes, can i activate a bflag on a running instance?
1396599171 M * daniel_hozac yes, use vattribute
1396599194 M * daniel_hozac note that NET_RAW will let your guest see all traffic on all guests and the host.
1396599819 J * fisted_ ~fisted@xdsl-81-173-190-177.netcologne.de
1396600109 M * jakobsg noted - thanks
1396600261 Q * fisted Ping timeout: 480 seconds
1396600261 N * fisted_ fisted
1396602328 Q * SteeleNivenson_ Ping timeout: 480 seconds
1396602331 Q * SteeleNivenson Ping timeout: 480 seconds
1396602601 J * jakobsg_ ~chatzilla@62.116.207.125
1396603023 Q * jakobsg Ping timeout: 480 seconds
1396604174 Q * jakobsg_ Quit: ChatZilla 0.9.90.1 [Firefox-Trunk 20.0a1/20130105061132]
1396604581 N * l0kit Guest5407
1396604587 J * l0kit ~1oxT@0001b54e.user.oftc.net
1396604976 Q * Guest5407 Ping timeout: 480 seconds
1396610049 J * thierryp_ ~thierry@zebra.inria.fr
1396610049 Q * thierryp Read error: Connection reset by peer
1396612802 Q * fisted Remote host closed the connection
1396612811 J * fisted ~fisted@xdsl-81-173-190-177.netcologne.de
1396613646 Q * thierryp_ Remote host closed the connection
1396614062 J * thierryp ~thierry@zebra.inria.fr
1396620124 J * thierryp_ ~thierry@zebra.inria.fr
1396620124 Q * thierryp Read error: Connection reset by peer
1396625900 Q * thierryp_ Remote host closed the connection
1396630523 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:e855:e029:9226:5732
1396631294 J * zerick ~eocrospom@190.187.21.53
1396632382 Q * fisted Ping timeout: 480 seconds
1396632999 J * SteeleNivenson ~SteeleNiv@pool-108-29-139-222.nycmny.fios.verizon.net
1396634973 N * Bertl_oO Bertl
1396634976 M * Bertl back now ...
1396640192 J * bonbons ~bonbons@2001:a18:209:4501:edc5:1114:8334:e015
1396648659 J * benl ~benl@dockoffice.sonassihosting.com
1396648664 M * benl Hey Bertl
1396648678 M * benl Just to let you know I traced the file descriptor leak back to Munin
1396648710 M * benl Looked like it began happening following an upgrade from 1.4 to 2.0
1396648737 M * benl I removed/purged then installed the packages again and have since been monitoring stats on the host and they've stayed stable
1396648770 M * benl But in terms of accounting for the "leaked" handles, that will remain an unknown, the only remedy thus far appears to be a host restart
1396648818 M * benl I made this rather beautiful Munin replacement to graph the usage
1396648827 M * benl See http://paste.linux-vserver.org/58644 (zoom out a little with your browser for correct view)
1396648966 M * benl And for anyone else, here's the script source (http://paste.linux-vserver.org/58645)
1396649014 M * Bertl so munin inside the guest, yes?
1396649049 M * benl Correct
1396649101 M * Bertl interesting
1396649119 M * benl Certainly is.
1396649137 M * benl I would love to find/release the "leaked" handles - but I've no idea where to begin
1396649157 M * benl So I'll just have to restart the machine in the coming weeks
1396649183 M * Bertl well, if you can kind of trigger it with that old plugin, I'd do exactly that with some instrumentation in place
1396649216 M * benl To be honest, the same 1.4>2.0 upgrade has run on about 150+ other machines without fault
1396649217 M * daniel_hozac so just to make sure i understand, it only happens if you have 1.4, then upgrade to 2.0? a fresh install of 2.0 doesn't see it?
1396649233 M * benl @Daniel - I've not exhaustively tested the hypothesis
1396649252 M * benl I basically stopped all daemons in the guest - and started them one at a time - observing file handles in the mean time
1396649275 M * benl The usage only started to escalate when I (manually) ran the munin-cron script
1396649293 M * daniel_hozac could this be linked to your broken hashify?
1396649298 M * daniel_hozac or was that on a different machine?
1396649302 M * benl Same machine.
1396649321 M * benl Could be related, but I'm not smart enough to say if it was or not!
1396649363 M * benl I've started testing of a 3.10 kernel now
1396649369 M * benl so I'll see how things go with that
1396649424 M * benl Out of interest, what's a recommended util-vserver release for 3.10?
1396649437 M * benl (with 2.3.6.8 patch)
1396649465 M * daniel_hozac latest is always recommended regardless of kernel.
1396649482 M * benl Latest stable, or latest kernel :)
1396649491 M * benl s/kernel/testing/g
1396649524 M * daniel_hozac testing. stable is not recommended for anything...
1396649529 M * benl lol
1396649615 M * benl I've not tried to replicate the 1.4>2.0 issue
1396649622 M * benl As I doubt I can
1396649638 M * benl I did it on hundreds of machines around the same time - and this is the only one that showed those symptoms
1396651553 J * fisted ~fisted@xdsl-84-44-239-137.netcologne.de
1396652757 Q * bonbons Quit: Leaving
1396652844 Q * benl Quit:  HydraIRC -> http://www.hydrairc.com <- The professional IRC Client :D
1396652865 Q * thierryp Remote host closed the connection
1396655623 J * undefined ~undefined@00011a48.user.oftc.net
1396655882 J * thierryp ~thierry@home.parmentelat.net