1345596045 J * bergerx_ ~bergerx@178.233.12.223 1345596483 Q * bergerx___ Ping timeout: 480 seconds 1345602252 Q * Romster Quit: Geeks shall inherit properties and methods of object earth. 1345602268 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com 1345602583 Q * Romster Quit: Geeks shall inherit properties and methods of object earth. 1345602599 J * Romster ~romster@202.168.100.149.dynamic.rev.eftel.com 1345603714 Q * Aiken Quit: Leaving 1345604715 J * bzed_ ~bzed@bzed.netrep.oftc.net 1345605024 Q * bzed Ping timeout: 480 seconds 1345605029 N * bzed_ bzed 1345605437 J * tam ~tam@says.screwallofyoubitches.com 1345605800 Q * nkukard Ping timeout: 480 seconds 1345605930 J * nkukard ~nkukard@197.87.42.133 1345606274 Q * clopez Ping timeout: 480 seconds 1345606285 Q * Chlorek Ping timeout: 480 seconds 1345606340 J * Chlorek chlorek@chlorek.com 1345606838 Q * tam Ping timeout: 480 seconds 1345606917 J * fisted_ ~fisted@xdsl-87-78-15-133.netcologne.de 1345607175 Q * fisted Ping timeout: 480 seconds 1345608241 J * tam ~tam@says.screwallofyoubitches.com 1345612767 J * kir ~kir@swsoft-msk-nat.sw.ru 1345612857 P * kir 1345613303 M * Bertl_oO off to bed now ... have a good one everyone! 1345613311 N * Bertl_oO Bertl_zZ 1345614023 Q * m_ueberall Remote host closed the connection 1345616210 J * ghislain ~AQUEOS@adsl2.aqueos.com 1345619446 Q * hparker Read error: Operation timed out 1345619470 J * hparker ~hparker@2001:470:1f0f:32c:beae:c5ff:fe01:b647 1345623064 J * vspas ~vspas@87.213.36.165 1345625882 Q * hparker Ping timeout: 480 seconds 1345626303 J * hparker ~hparker@2001:470:1f0f:32c:beae:c5ff:fe01:b647 1345627672 J * ensc ~irc-ensc@p54ADF8E7.dip.t-dialin.net 1345629917 J * clopez ~clopez@fanzine.igalia.com 1345630061 M * disposable i have a SNAFU situation. i was running a vserver stored on an NFS share. that NFS server died this morning and now i'm stuck with a vserver that i can't stop. is there a way to kill such a vserver? 1345630307 M * Jb_boin have you tried vkill -s 9 --xid VSERVERXID 1345631024 M * disposable Jb_boin: i have not. since there is no manual page for it, i have no idea what it does 1345631105 M * daniel_hozac it might be related to killing something 1345631288 M * disposable daniel_hozac: very funny. let me rephrase that - since there's no manual page for it, 'apropos vserver' doesn't return that command's name AND without a manual page i would still assume it only kills processes, not the actual vserver too. 1345631333 M * disposable now that i know of vkill's existence, google has provided useful links 1345631338 M * disposable Jb_boin: thanks 1345631418 M * daniel_hozac yes, it only kills processes. 1345631423 M * daniel_hozac a guest dies when the last process in it dies. 1345631431 M * daniel_hozac (unless it is persistent) 1345631442 M * disposable daniel_hozac: that's good to know. thanks 1345632206 M * Jb_boin :) 1345632230 M * Jb_boin its used to kill guests processes from the host 1345632243 M * disposable ok, mine's the persistent kind. according to vps, it is still running two processes. (rsyslogd and unbound) any tips on how to kill them? (vkill --xid 1075 -s 9 did nothing) 1345632268 M * Jb_boin if you dont specify a process it will kill them all (and you dont need to specify xid if you specify process(es)) 1345632269 M * daniel_hozac that doesn't mean it's persistent. 1345632283 M * daniel_hozac what state are they in? 1345632289 M * Jb_boin they must be in D state 1345632318 M * Jb_boin on some kernels... you cant kill some processes and have to reboot to be able to restart the vserver 1345632340 M * disposable this is the output of my vps http://pastebin.com/DZDBC6S4 1345632362 N * Bertl_zZ Bertl 1345632367 M * Bertl morning folks! 1345632433 M * Jb_boin you can check parent process with vps alx 1345632464 M * Bertl might be a side effect of the reaper changes? 1345632476 M * daniel_hozac they're just D-processes. 1345632483 M * daniel_hozac probably waiting on NFS. 1345632545 M * disposable yes, rpc http://pastebin.com/S86HuAWc 1345632568 M * Bertl ah, yes, just read up on the logs, server gone, probably mounted soft 1345632626 M * disposable /etc/vservers/dns4/vdir was a softlink to a dir inside and nfs mount 1345632629 M * disposable s/and/an 1345632643 M * Bertl what does /proc/mounts show regarding the nfs mount? 1345632835 M * disposable Bertl: nothing, it's no longer in /proc/mounts. nfs server had problems, so i stopped all vservers that were running off it (forgot about this one), unmounted the share and shut down the nfs server. a few hours later i noticed this vserver having problems. 1345632867 M * Bertl check inside the guest namespace 1345632906 M * Bertl e.g. do vnamespace -e -i 0 -- /bin/bash 1345632993 M * disposable unknown option `-i' 1345633129 M * disposable Bertl: when i ran it without "-i 0" i got a carriage return and no bash prompt. i can't quit it now 1345633266 M * disposable Bertl: had to disconnect and reconnect to base server 1345633467 M * Bertl update your uti-vserver to something recent :) 1345633637 M * disposable Bertl: I can't do that. it's a live server. it is using whatever came with ubuntu 10.04 (and running christopher lukas's 2.6.32-28-vserver kernel from ppa recommented on linux-vserver website) 1345634002 M * Bertl well, that's up to you, but judging from the missing -i, it might easily be that the tools are too old to achieve proper guest isolation and your host system can easily be compromised 1345634110 M * disposable we are slowly moving everything to newer systems so this will be addressed. any clues as to what to do in this situation? 1345634174 M * Bertl depends on how the nfs was mounted 1345634193 M * Bertl in general the best solution is to bring the nfs server back up 1345634203 M * disposable vers=3,async,noatime,users,noauto,exec 1345634247 M * disposable this is possible but something i was trying to avoid. (a few hours of work) 1345634698 M * Bertl for that kernel, without specifying anything about the mount mode (hard/soft) your nfs mount is probably hard,timeo=600 1345634715 M * Bertl Hard 1345634716 M * Bertl retries indefinitely. The client will continue to attempt the NFS file operation indefinitely if the operation fails. Hard mounts should be used when the server and the link to the server are known to be reliable. Hard mounts with the interruptible option enabled is the recommended method of mounting remote filesystems. The default is hard mount. 1345635219 Q * hparker Ping timeout: 480 seconds 1345636058 M * disposable Bertl: thank you, i put the dead server's drives into another server, changed a few udev rules and as soon as i mounted the nfs share, the vserver immediately shut down. 1345636233 M * Bertl you're welcome! 1345636259 M * Bertl next time, mount it soft if the server is unreliable 1345636282 Q * fisted_ Remote host closed the connection 1345636285 J * fisted ~fisted@xdsl-87-78-15-133.netcologne.de 1345636904 M * disposable Bertl: it is reliable until it breaks 1345638597 J * ghislain1 ~AQUEOS@adsl2.aqueos.com 1345638892 Q * ghislain Ping timeout: 480 seconds 1345640744 Q * nox Ping timeout: 480 seconds 1345641290 J * nox ~bnc@2a01:4f8:140:40a2:6667::952 1345641951 M * Bertl translocating ... bbl 1345641955 N * Bertl Bertl_oO 1345643819 Q * clopez Ping timeout: 480 seconds 1345643861 M * aurel_ Meh. I moved this application from a vserver on a slow server to a vserver on a (allegedly) fast server, just to see it develop different kinds of performance and network issues. Weird. 1345645225 M * Jb_boin can you be more explicit about the issues? :) 1345645877 J * clopez ~clopez@fanzine.igalia.com 1345646012 J * BlackPanx 4d6f0224@ircip2.mibbit.com 1345646018 M * BlackPanx hey everyone 1345646047 M * BlackPanx a question i have. 1345646071 M * BlackPanx i'm trying to set up serial console between two servers. I exported and started getty on ttyS0 1345646092 M * BlackPanx i am able to get data from that server to the other server with minicom program 1345646114 M * BlackPanx the reason i made this is to catch kernel panic issues on one of my servers 1345646132 M * BlackPanx will server send kernel panic to ttyS0 as well as it does to tty1 1345646138 M * BlackPanx or do i have to do something else 1345646151 M * daniel_hozac did you boot with console=ttyS0? 1345646161 M * BlackPanx do i need to ? 1345646182 M * BlackPanx can i have tty1 AND ttyS0 1345646184 M * BlackPanx both 1345646184 M * BlackPanx ? 1345646534 M * BlackPanx linux /vmlinuz-2.6.38.8-vs2.3.0.37-rc17 root=UUID=8f5a938a-6f4d-4aba-9f9f-986a4b1a5c34 ro console=ttyS0,115200n8 console=tty0 quiet splash 1345646538 M * BlackPanx should work right ? 1345646541 M * BlackPanx for example 1345646560 M * daniel_hozac probably 1345646852 M * aurel_ Jb_boin: sorry, I think the unspecific issues were symptoms of a disk i/o bottleneck, i/o stalling because the system was writing large files during a raid resync. I'll solve it by throwing hardware at it. 1345646984 M * Jb_boin BlackPanx, well, 2.6.38 were prone to crashs in the days and if your kernel is prtty old it might ve the same issue i ran into 1345647002 M * Jb_boin had this from 2.6.36.x to 3.0.x 1345647018 M * Jb_boin latest 3.0.x works fine and i never had an issue with 3.2.x 1345647155 M * BlackPanx Jb_boin 3.2 with vserver patch you mean ? 1345647174 M * BlackPanx i keep having kernel panics at times of congestion 1345647192 M * BlackPanx meaning if server works a lot, it's high probability it'll crash 1345647197 M * Jb_boin yep 1345647197 M * BlackPanx with kernel panic or something 1345647213 M * Jb_boin had that, caused by high apache load most of the time 1345647236 M * Jb_boin were having ~1 kernel panic/day over ~40 servers 1345647241 M * BlackPanx vmlinuz-3.0.24-vs2.3.2.3-black i have compiled this one 1345647263 M * Jb_boin if you are on debian/ubuntu you can go with beng precompiled kernels 1345647272 M * Jb_boin BlackPanx, i would even recommand 3.2.x 1345647284 M * BlackPanx you use your own kernel or from distro ? 1345647287 M * BlackPanx from repo 1345647289 M * Jb_boin but 3.0 should also be fine 1345647290 M * Jb_boin from repo 1345647293 P * tam 1345647308 M * BlackPanx you tested 3.0 one right ? 1345647317 M * Jb_boin yep, had some for months 1345647331 M * Jb_boin but dont have 3.0.x anymore i think 1345647337 M * BlackPanx well the thing is i just want the crashes to stop. we have like 8 physical servers... and when high load they tend to crash from time to time 1345647347 M * Jb_boin at least not on the web cluster 1345647348 M * BlackPanx and like 200 virtual servers 1345647378 M * Jb_boin we got 3.2 on ~50-60 servers 1345647384 M * Jb_boin and no great issues so far 1345647423 M * BlackPanx this sounds perfect for us 1345647429 M * BlackPanx as long as crashes stop 1345647441 M * BlackPanx cause we do not have any problems so far except these annoying crashes 1345647444 M * BlackPanx that are totally random 1345647446 M * BlackPanx -.- 1345647453 M * BlackPanx thanx for the tip. 1345647465 M * BlackPanx would you mind telling me repos you use for vserver kernel ? 1345647480 M * BlackPanx we have ubuntu and centos mainly 1345647496 M * BlackPanx lucid 1345647504 M * BlackPanx 10.04 1345647522 M * BlackPanx and CentOS 6.0 1345647531 M * Jb_boin http://repo.psand.net/ 1345647535 M * Jb_boin its for debian/ubuntu 1345647547 M * Jb_boin deb http://repo.psand.net/ squeeze main 1345647568 M * Jb_boin but on debian, initramfs wont generate automatically when you install a new kernel and thus, kernel wont boot 1345647587 M * BlackPanx i know 1345647592 M * BlackPanx you need to start it manually 1345647592 M * Jb_boin in order to do that you must do "update-initramfs -k 3.2.23-vs2.3.2.12-beng -c ; update-grub" 1345647593 M * BlackPanx to generate 1345647596 M * BlackPanx the initramfs 1345647598 M * BlackPanx true 1345647599 M * BlackPanx yes 1345647604 M * Jb_boin while its working on debian 1345647621 M * Jb_boin its always better to know before than having a non-booting kernel :) 1345647630 M * Jb_boin and not knowing why 1345647738 M * BlackPanx true 1345647739 M * BlackPanx :) 1345647742 M * BlackPanx thanks big time 1345647854 M * BlackPanx did you have any issues with vservers upgrading from 2.6 to 3.0 / 3.2 ? 1345647870 M * BlackPanx updating util-vserver of course also. 1345648994 Q * vspas Ping timeout: 480 seconds 1345649149 J * ryker ~Adium@c-67-176-243-86.hsd1.in.comcast.net 1345649867 J * m_ueberall dircproxy4@2a01:4f8:100:80e3:0:bc28:af65:3 1345650703 J * fisted_ ~fisted@xdsl-78-35-81-214.netcologne.de 1345650893 Q * fisted Ping timeout: 480 seconds 1345651597 J * bonbons ~bonbons@2001:960:7ab:0:e074:5a98:aecc:58b0 1345651626 M * BlackPanx do we have any more recent repository for ubuntu vserver ? anyone knows about it ? 1345652290 N * Bertl_oO Bertl_zZ 1345653065 M * daniel_hozac beng's repository should be fine, i imagine. 1345653079 M * BlackPanx ye but that's for debian 1345653102 M * BlackPanx on ubuntu and debian they strongly suggest not to use debian precompiled kernels on ubuntu 1345653116 M * BlackPanx even tho the distro should be same "squeeze/sid" in this case 1345653128 M * BlackPanx i guess i could give it a try on dev server maybe 1345653131 M * daniel_hozac it's a vanilla kernel, i don't think there'd be any noticable difference. 1345653151 M * BlackPanx u think it shouldn't make any difference ? 1345653254 M * BlackPanx i'll try it 1345653256 M * BlackPanx maybe it works. 1345653262 M * BlackPanx would be neat tho 1345655488 Q * bonbons Ping timeout: 480 seconds 1345656140 M * Jb_boin BlackPanx, didnt have any issues upgrading util-vserver except that sometimes you need to install in 2 times the package 1345656163 M * Jb_boin like apt-get install util-vserver will fail but if you do a apt-get -f install afterwise it will finish the install 1345656182 M * Jb_boin and beng packages works fine on ubuntu 10.04 1345656187 M * Jb_boin didnt test on 12.04 so far 1345656265 J * bonbons ~bonbons@2001:960:7ab:0:e074:5a98:aecc:58b0 1345656776 Q * ensc|w Remote host closed the connection 1345656836 Q * FireEgl Ping timeout: 480 seconds 1345656929 Q * m_ueberall Remote host closed the connection 1345662674 Q * clopez Ping timeout: 480 seconds 1345663291 J * m_ueberall dircproxy4@2a01:4f8:100:80e3:0:bc28:af65:3 1345663923 Q * imcsk8 Remote host closed the connection 1345663923 Q * ichavero_ Remote host closed the connection 1345663924 Q * ichavero__ Read error: Connection reset by peer 1345667307 Q * bonbons Quit: Leaving 1345667797 N * Bertl_zZ Bertl 1345667802 M * Bertl back now ... 1345671120 Q * ntrs Ping timeout: 480 seconds 1345671475 J * ntrs ~ntrs@vault08.rosehosting.com 1345672256 J * clopez ~clopez@65.27.165.83.dynamic.mundo-r.com 1345672270 N * ensc Guest3962 1345672280 J * ensc ~irc-ensc@p4FFCFD8D.dip.t-dialin.net 1345672688 Q * Guest3962 Ping timeout: 480 seconds 1345674104 Q * ghislain1 Quit: Leaving. 1345678206 J * agaran ~agaran@static-78-8-120-176.ssp.dialog.net.pl 1345678217 M * agaran hello 1345678246 Q * PowerKe Ping timeout: 480 seconds 1345678579 M * agaran is there some easy way to find which vserver got bound to specific ip/port (ie, on which of vservers lighttpd grab :80) ? 1345679016 J * PeteG_ ~vampirnat@178.114.243.44.wireless.dyn.drei.com 1345679032 M * nlm uhm, I suppose you can run netstat on the monitor context and check all LISTEN on :80 from all contexts 1345679059 M * agaran plain netstat --inet --inet6 -nap don't show anything but main system.. 1345679069 M * agaran otherwise i wouldnt ask here :-) 1345679088 M * nlm I meant to run netstat on the monitor context, (context id 1) 1345679095 M * nlm like vtop does 1345679097 J * PowerKe ~tom@94-226-105-17.access.telenet.be 1345679124 Q * PeteG Ping timeout: 480 seconds 1345679138 M * nlm something like: chcontext --silent --ctx 1 netstat etcetcetc 1345679224 M * agaran i am affraid that dont show anything usefull, shows main system processes only 1345679250 M * agaran i wanted to avoid downing all vservers just to find which one causes problem thats all 1345679297 N * Bertl Bertl_oO 1345679317 M * nlm ok, sorry. I'm kind of new on this but it worth the try :) 1345679333 M * agaran nlm: its usefull command, of course worth to try :) 1345679345 M * agaran i didnt knew that ctx 1 has some special meaning 1345679423 M * nlm it's a context that it's supposed to see all processes from all contexts 1345679437 M * nlm I thought it was the same about networking, but it's not 1345679468 M * agaran found problematic one seems one of earlier configured vservers grabbed lastly added ip somehow, so i messed config somewhere 1345679838 Q * ryker Quit: Leaving.