1452729809 Q * Wermwud Quit: Going offline, see ya! (www.adiirc.com)
1452732273 J * clopez ~tau@neutrino.es
1452733143 Q * fstd Remote host closed the connection
1452733155 J * fstd ~fstd@xdsl-87-78-8-213.netcologne.de
1452739496 Q * derjohn_mob Ping timeout: 480 seconds
1452740037 N * Bertl_zZ Bertl
1452740039 M * Bertl back now ...
1452740065 M * Bertl Guy-: can you upload the dump?
1452740146 M * Bertl I do not know any NICs which could do 30k frames
1452743163 J * thierryp ~thierry@2a01:e35:2e2b:e2c0:3d33:79cb:2c67:c299
1452743164 Q * thierryp_ Read error: Connection reset by peer
1452751901 Q * thierryp Remote host closed the connection
1452756129 J * Ghislain ~aqueos@adsl1.aqueos.com
1452756656 J * thierryp ~thierry@zeta.inria.fr
1452757971 J * derjohn_mob ~aj@p578b6aa1.dip0.t-ipconnect.de
1452761748 M * Guy- yes, I can upload dumps
1452761783 M * Guy- it certainly seems tcpdump doesn't show me what's actually on the wire; I obtained dumps of the same streams on all four participating interfaces and they look radically different
1452762319 M * Guy- I'll repeat the exercise with _all_ kinds of offloading disabled that even can be disabled
1452762539 M * Guy- Bertl: this is a base64 encoded tar.bz2 of the dump files: http://sprunge.us/FNga
1452762558 M * Guy- there is even a purported ~60k frame in there
1452762570 M * Guy- but this _must_ be some sort of artifact
1452763015 Q * derjohn_mob Ping timeout: 480 seconds
1452763512 M * Bertl looks like you have some offloading active in the kernel/driver, yes?
1452763544 M * Bertl check with ethtool -k
1452763628 M * Bertl I'd expect *segmentation-offload and maybe large-receive-offload
1452764481 J * derjohn_mob ~aj@fw.gkh-setu.de
1452767169 M * Bertl off for now ... bbl
1452767170 N * Bertl Bertl_oO
1452770580 M * Guy- Bertl_oO: yes, I suspected those myself and disabled them (on the sending system at least)
1452770593 M * Guy- it only has scatter-gather offloading enabled now
1452770604 M * Guy- but I'll see what happens if I disable all offloading everywhere
1452772730 M * Guy- OK, I'm stupid; it's offloading all right, and while I did disable it, I disabled it on the wrong interface.
1452772889 M * Guy- tcp-segmentation-offload, to be precise
1452776343 Q * fstd Remote host closed the connection
1452776355 J * fstd ~fstd@xdsl-87-78-143-222.netcologne.de
1452780606 J * Ghislain1 ~aqueos@adsl1.aqueos.com
1452780792 Q * Ghislain Ping timeout: 480 seconds
1452786361 M * Bertl_oO yeah, that was what I suspected :)
1452786368 N * Bertl_oO Bertl
1452788458 Q * thierryp Remote host closed the connection
1452789639 M * Guy- damn interfaces look so alike :)
1452790216 M * Bertl you know, you can rename them :)
1452790273 Q * derjohn_mob Ping timeout: 480 seconds
1452790296 Q * Ghislain1 Ping timeout: 480 seconds
1452792805 J * derjohn_mob ~aj@tmo-100-233.customers.d1-online.com
1452797004 Q * derjohn_mob Ping timeout: 480 seconds
1452797919 P * sannes 
1452798155 J * sannes ~ace@2a02:fe0:c131:9070:558b:62df:a116:37c1
1452801244 M * Guy- hah! Real Men (tm) remember which one is eth4 and which is eth5! ... eventually.
1452801529 J * derjohn_mob ~aj@tmo-100-170.customers.d1-online.com
1452802161 M * Bertl Real Men (tm) don't complain about the interface names *G*
1452804232 Q * karasz Remote host closed the connection
1452804237 J * karasz ~karasz@00015555.user.oftc.net
1452804277 Q * funnel Remote host closed the connection
1452804287 J * funnel ~funnel@0001c7d4.user.oftc.net
1452804458 Q * a1-away Ping timeout: 480 seconds
1452805022 J * a1-away ~jelle@62.27.85.48
1452805165 M * gnarface hey i'm still running 3.14.17-vs2.3.6.13
1452805168 M * gnarface is that a good idea?
1452805191 M * gnarface or is there a better patch/kernel i should be using now?
1452805276 M * gnarface (i'm asking because the server apparently fully I/O locked on me while i was sleeping yesterday, after +300 days of uptime, so i'm trying to figure out whether it was a software or hardware fault)
1452809582 M * gnarface if its *not* a hardware fault, the primary culprits are apache2, a munin plugin, or one of the maps i added to the ioquake3 server
1452809594 M * gnarface primary suspects, i mean
1452809619 M * gnarface because nothing else really had changed on the server in that +300 days of uptime
1452811011 M * Bertl well, for sure some bugs and probably exploits have been discovered and fixed since it was released
1452811052 M * Bertl but usually Linux-VServer systems are somewhat immune because of the reduced capabilities
1452811118 M * Bertl we have a patch for 3.14.52 online, which probably applies to a later 3.14.x version as well
1452814365 M * undefined gnarface, Bertl: i've successfully patched, compiled, and tested 3.14.58-vs2.3.6.15