1216339209 Q * balbir Ping timeout: 480 seconds 1216339475 Q * docelic Quit: http://www.spinlocksolutions.com/ 1216340745 J * esa` ~esa@ip-87-238-2-45.static.adsl.cheapnet.it 1216340749 Q * esa Ping timeout: 480 seconds 1216340856 J * doener ~doener@i577BAA6E.versanet.de 1216340962 Q * doener_ Ping timeout: 480 seconds 1216346086 J * nenolod_ ~nenolod@ip70-189-74-62.ok.ok.cox.net 1216346103 Q * nenolod_ Remote host closed the connection 1216346525 J * nenolod_ ~nenolod@ip70-189-74-62.ok.ok.cox.net 1216346558 Q * nenolod_ Remote host closed the connection 1216347424 J * Moser__ ~chatzilla@Zfa27.z.pppool.de 1216347425 N * Moser__ Moser 1216347761 Q * Moser_ Ping timeout: 480 seconds 1216347851 J * nenolod_ ~nenolod@ip70-189-74-62.ok.ok.cox.net 1216347862 Q * nenolod_ 1216348676 Q * nenolod Quit: nenolod: why do you automatically assume I want friends? 1216348687 J * nenolod ~nenolod@ip70-189-74-62.ok.ok.cox.net 1216349557 J * balbir ~balbir@12.157.249.2 1216354628 Q * FloodServ synthon.oftc.net services.oftc.net 1216358262 J * ruskie ruskie@goatse.co.uk 1216358608 J * derjohn_mob ~aj@p5B23DC30.dip.t-dialin.net 1216359440 Q * derjohn_mob Ping timeout: 480 seconds 1216361642 J * ntrs ~ntrs@77.29.64.224 1216361862 J * cryptronic ~oli@p54A3B4DB.dip0.t-ipconnect.de 1216362763 M * yang I am wondering how could I mount this fstab line on my guest without rebooting "/mnt/disk2 /mnt/storage none bind 0 0" 1216362803 M * yang it's the entry from /etc/vservers//fstab 1216362833 M * cehteh hehe .. seems to be a common problem 1216362836 M * cehteh [14:06] hrm, so is there a way to remount that /etc/vserver/*/fstab mount without rebooting the vserver? 1216362836 M * cehteh [14:09] yes .. but quite intrusive/manual way 1216362836 M * cehteh [14:10] hrm, annoying 1216362836 M * cehteh [14:10] write a tool for it .. vserver $name remount ... 1216362836 M * cehteh [14:10] i would like it :) 1216362838 M * yang really? 1216362843 M * cehteh .. from yesterday :) 1216362867 M * yang ok, i will reboot then :/ 1216362906 M * cehteh restart the vserver, not full reboot 1216362909 M * yang yep 1216362930 M * cehteh well you can do it manually .. just figure out how the vserver start scripts do it ... 1216362940 M * cehteh changing namespace and so on 1216362964 M * cehteh it shouldnt be too hard 1216362995 M * yang what do you mean? 1216363002 M * cehteh while a vserver $something mount would be really the way to go 1216363052 M * yang its kind of a problem to restart the guest, becouse there are many processes running 1216363075 M * yang isn't there some option with mount --bind etc. 1216363114 J * z0d ~z0d@fw.wonderline.hu 1216363118 M * z0d re 1216363133 M * cehteh eh? 1216363159 M * cehteh the problem is rather that vserver mounts are done in their respective namespaces 1216363162 M * cehteh thats all 1216363242 M * yang well this restart isn't really needed now, I would like to preserve the uptime 1216363268 M * yang It will mount automatically on the next reboot 1216364471 M * arachnist matti: http://www.joemonster.org/i/a/chinski.gif 1216364887 J * FloodServ services@services.oftc.net 1216367123 J * kir ~kir@swsoft-msk-nat.sw.ru 1216367551 N * DoberMann[ZZZzzz] DoberMann 1216368549 J * pisco ~pisco@tor.noreply.org 1216369299 Q * ntrs Ping timeout: 480 seconds 1216369535 J * ema ~ema@rtfm.galliera.it 1216369860 Q * kir Remote host closed the connection 1216370312 Q * hparker Quit: Read error: 104 (Peer reset by connection) 1216375968 J * mrjack ~mrjack@office.smart-weblications.net 1216375969 M * mrjack hi 1216376834 Q * cryptronic Quit: Leaving. 1216377929 J * ntrs ~ntrs@77.29.73.226 1216378221 M * mrjack finally.. 1216378240 M * mrjack converted appr. 300 vservers from 1.x to 2.x config style and all are up and running... 1216378386 J * pisco_ ~pisco@tor.noreply.org 1216378394 Q * pisco Ping timeout: 480 seconds 1216380205 Q * ntrs Ping timeout: 480 seconds 1216381110 J * docelic ~docelic@78.134.204.53 1216383357 J * docelic_ ~docelic@78.134.206.58 1216383566 M * sid3windr all with that small cript heh 1216383568 M * sid3windr +s 1216383573 M * sid3windr I can imagine you didn't want to do that by hand 1216383760 Q * docelic Ping timeout: 480 seconds 1216383860 J * kir ~kir@swsoft-msk-nat.sw.ru 1216383862 Q * kir Remote host closed the connection 1216383866 J * kir ~kir@swsoft-msk-nat.sw.ru 1216383866 Q * kir Remote host closed the connection 1216383872 J * kir ~kir@swsoft-msk-nat.sw.ru 1216383873 Q * kir Remote host closed the connection 1216383881 J * kir ~kir@swsoft-msk-nat.sw.ru 1216383881 Q * kir Remote host closed the connection 1216383888 J * kir ~kir@swsoft-msk-nat.sw.ru 1216383889 Q * kir Remote host closed the connection 1216383892 N * Bertl_zZ Bertl 1216383921 M * Bertl morning folks! 1216383932 J * kir ~kir@swsoft-msk-nat.sw.ru 1216383933 Q * kir Remote host closed the connection 1216383978 J * kir ~kir@swsoft-msk-nat.sw.ru 1216383996 M * Bertl kir: please fix your client 1216384010 P * kir 1216384942 M * sid3windr :p 1216385717 J * Slydder_ ~slydder@194.59.17.53 1216385745 Q * Slydder_ 1216385916 Q * Aiken Quit: Leaving 1216386092 Q * pisco_ Remote host closed the connection 1216386134 J * pisco ~pisco@tor.noreply.org 1216386187 M * daniel_hozac yang, cehteh: vmount is supposed to be able to do it. 1216386258 M * daniel_hozac Bertl: http://people.linux-vserver.org/~dhozac/p/k/delta-vspid-revert01.diff 1216386269 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-pidspace-feat04.diff 1216386274 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-pidspace-lock04.diff 1216386279 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-pidns-fix01.diff 1216386281 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-pidns-fix02.diff 1216386284 M * daniel_hozac http://people.linux-vserver.org/~dhozac/p/k/delta-pidns-fix03.diff 1216386844 M * Bertl nice, do we need the get_proc_task() at all? 1216386890 M * Bertl or is that in mainline now? 1216386978 M * daniel_hozac hmm? get_proc_task is mainline. 1216386997 M * Bertl okay, then it's fine, I thought we added that at some point 1216387041 M * Bertl we want to remove the reaper comment in context.c too, I guess 1216387054 M * Bertl (but you reuse that later, I presume?) 1216387109 M * daniel_hozac yes. 1216387219 M * Bertl okay, that is not against the recent patch, right? 1216387241 M * Bertl (looking at the CLONE_NEWPID, which I think I left commented out) 1216387247 M * daniel_hozac no, 13.1 was all i had. 1216387256 M * Bertl okay, no problem 1216387314 M * Bertl okay, we have to talk about the lock patch .. first I think we should not make wrappers for the locking primitives, that just confuses the reader 1216387371 M * yang daniel_hozac: vmount? 1216387416 M * Bertl daniel_hozac: the second thing is the type of lock, basically we can choose between mutex, rwlock and spinlock 1216387429 M * Bertl (depending on the usage profile) 1216387443 Q * balbir Ping timeout: 480 seconds 1216387503 J * ntrs ~ntrs@77.29.73.226 1216387559 M * Bertl daniel_hozac: what was the locking issue you saw, besides the ref-count dropping on proxy/fs? (which would be fixed by taking a reference) 1216387701 M * daniel_hozac dereferencing and increasing the ref-count is not atomic. 1216387764 M * Bertl well, a get() function is sufficient 1216387803 M * Bertl i.e. increment + read (atomic) then check 1216387894 M * daniel_hozac vx_initpid and such are what i'm most worried about, where we dereference pointers that might've canged 4 times. 1216387943 M * Bertl well, it's not a performance critical path, so I'm fine with locking 1216387953 M * daniel_hozac none of them are. 1216387955 M * Bertl what was the reason for rw locks= 1216387974 M * daniel_hozac it's mostly read. 1216388003 M * daniel_hozac and concurrent readers are fine. 1216388046 M * Bertl okay, should be fine then, but we should replace the wrappers with direct locking primitives for readability 1216388123 M * Bertl the proc path stuff was worked around? 1216388135 M * Bertl (i.e. I didn't see it right now :) 1216388194 M * daniel_hozac proc path stuff? 1216388206 M * Bertl where you changed a kmalloc to a stack variable 1216388235 M * Bertl (with MAX_PATHLEN >> 2 or so) 1216388237 J * yarihm ~yarihm@whitehead2.nine.ch 1216388277 M * daniel_hozac hmm, you're right, i didn't merge that hunk. 1216388477 M * Bertl the pidns-fix02 I still consider wrong .. ifthat is only checked for existance somewhere, we should remove that check, not fake the entry 1216388762 M * daniel_hozac PROC_I(inode)->pid has to be set to something. 1216389620 Q * z0d Remote host closed the connection 1216389915 M * Bertl okay, will look into that a little later .. have to run now ... 1216389921 N * Bertl Bertl_oO 1216390222 J * balbir ~balbir@bi01p1.co.us.ibm.com 1216391321 Q * Moser Remote host closed the connection 1216392181 J * fatgoose ~samuel@82.80.modemcable.oricom.ca 1216392804 J * fatgoose_ ~samuel@82.80.modemcable.oricom.ca 1216392810 Q * fatgoose_ 1216393133 Q * fatgoose Ping timeout: 480 seconds 1216393186 J * rgl ~rgl@bl8-129-150.dsl.telepac.pt 1216393190 M * rgl hi 1216393195 M * daniel_hozac hello 1216393320 J * fatgoose ~samuel@82.80.modemcable.oricom.ca 1216393439 M * rgl ever seen a SMART disk error? 1216393447 M * rgl mine seems to have some of them :( 1216393475 M * rgl like, the smart log has 33 of these: READ FPDMA QUEUED 1216393504 Q * ema Quit: leaving 1216395004 Q * balbir Ping timeout: 480 seconds 1216395060 J * er ~sapanbhat@aegis.CS.Princeton.EDU 1216395728 J * balbir ~balbir@bi01p1.co.us.ibm.com 1216395849 Q * eyck Ping timeout: 480 seconds 1216395953 Q * fatgoose Ping timeout: 480 seconds 1216396095 Q * pisco Remote host closed the connection 1216396145 J * fatgoose ~samuel@82.80.modemcable.oricom.ca 1216396268 J * rgl_ ~rgl@bl8-129-150.dsl.telepac.pt 1216396681 Q * rgl Ping timeout: 480 seconds 1216397438 J * eyck mhdCrUSh@nat03.nowanet.pl 1216397463 J * rgl ~rgl@bl8-134-113.dsl.telepac.pt 1216397554 J * Linus ~Nuhx@bl7-155-212.dsl.telepac.pt 1216397680 Q * rgl_ Ping timeout: 480 seconds 1216398106 J * bonbons ~bonbons@2001:960:7ab:0:2c0:9fff:fe2d:39d 1216398293 J * rgl_ ~rgl@bl8-134-113.dsl.telepac.pt 1216398403 Q * nkukard Remote host closed the connection 1216398453 J * nkukard ~nkukard@196.212.73.74 1216398705 Q * rgl Ping timeout: 480 seconds 1216399241 J * ntrs_ ~ntrs@77.29.67.65 1216399460 Q * rgl_ Quit: Leaving 1216399609 Q * ntrs Ping timeout: 480 seconds 1216399654 Q * yarihm Quit: This computer has gone to sleep 1216399987 Q * balbir Ping timeout: 480 seconds 1216400131 J * Moser ~chatzilla@Zfa27.z.pppool.de 1216400614 N * DoberMann DoberMann[PullA] 1216400983 J * yarihm ~yarihm@217.150.254.94 1216402041 J * Linus_ ~Nuhx@bl7-151-42.dsl.telepac.pt 1216402442 Q * Linus Ping timeout: 480 seconds 1216403146 J * balbir ~balbir@bi01p1.co.us.ibm.com 1216403265 Q * er Quit: er 1216404218 Q * yarihm Quit: This computer has gone to sleep 1216405329 Q * docelic_ Quit: http://www.spinlocksolutions.com/ 1216405674 M * snooze what is the actual difference between vserver stop and shutdown -h now inside the guest? 1216405701 M * daniel_hozac plain init? 1216405763 M * snooze yes 1216405817 M * daniel_hozac vserver stop does kill -s 1 1 1216405875 M * snooze and shutdown -h now doesnt? 1216405898 M * snooze because im getting timeout when trying to use stop for some reason 1216405975 M * daniel_hozac make sure your init is setup to handle ctrl-alt-del correctly. 1216405985 M * snooze oh 1216405994 M * snooze i actually removed that from inittab recently 1216406084 N * Linus_ Linus 1216406163 J * hparker ~hparker@linux.homershut.net 1216406299 M * snooze thanks, worked better now 1216407009 J * er ~sapanbhat@aegis.CS.Princeton.EDU 1216407324 J * yarihm ~yarihm@84-74-147-84.dclient.hispeed.ch 1216408164 J * lilalinux ~plasma@80.69.41.3 1216408594 Q * lilalinux Remote host closed the connection 1216409993 J * rgl ~rgl@lx2-84-90-11-153.netvisao.pt 1216409995 A * rgl waves 1216410043 Q * rgl Read error: Connection reset by peer 1216410293 Q * yarihm Quit: This computer has gone to sleep 1216410532 J * docelic ~docelic@78.134.206.58 1216410691 J * rgl ~rgl@lx2-84-90-11-153.netvisao.pt 1216412163 Q * Linus Quit: leaving 1216412343 J * dniel ~ary@host166.190-138-96.telecom.net.ar 1216412484 M * dniel I did the hashify like this site: http://linux-vserver.org/Frequently_Asked_Questions#How_do_I_manage_a_multi-guest_setup_with_vhashify.3F 1216412487 M * dniel Now how can I make new guest with unification? 1216412599 M * dniel vserver build -m clone --hostname name.mydomain.com --interface eth0:10.0.0.118/24 -- --source var/lib/vservers/ 1216412609 M * dniel is it correct? 1216412708 M * dniel It only made a new guest without unification :( 1216412726 M * dniel what is wrong? 1216413730 Q * er Quit: er 1216414052 Q * mrjack Ping timeout: 480 seconds 1216414056 J * er ~sapanbhat@aegis.CS.Princeton.EDU 1216414382 M * er i think I found the problem that was causing tcp payloads to get lost from in tcpdump 1216414403 M * er it's not a vserver problem, though 1216415943 Q * balbir Ping timeout: 480 seconds 1216416627 J * RenOS ~renos@athedsl-242152.home.otenet.gr 1216416630 M * RenOS hello all 1216416665 M * RenOS is there any good interface to manage the vserver contexts? 1216416670 M * RenOS (besides the cli) 1216417468 M * dniel RenOS: http://www.openvcp.org/ 1216417483 M * RenOS I've seen it but it seems it's dead 1216417490 M * RenOS no updates for the last year.. 1216417524 M * dniel RenOS: mmm yes :( 1216417526 M * dniel http://www.openvcp.org/ 1216417529 M * dniel sorry 1216417541 M * dniel http://www.openqrm.org/ 1216417570 M * RenOS openqrm doesn't support vserver. it's for openvz and xen 1216417583 M * RenOS as far I could tell by checking the web site out 1216417637 M * dniel in this interview bertl said this: http://www.montanalinux.org/linux-vserver-interview.html 1216417701 Q * er Quit: er 1216417717 M * RenOS yes, you are right... 1216417742 M * dniel RenOS: I never probed one.... but maybe it will work 1216417762 M * RenOS i was also looking as hypervm but still no vserver support 1216417788 M * dniel :( 1216417844 J * er ~sapanbhat@aegis.CS.Princeton.EDU 1216417865 M * RenOS is anyone here doing VPS business with vserver? 1216417972 M * dniel RenOS: http://www.openqrm.org/news/qlusters-names-new-executive-team-members-to-drive-sales-and-accelerate-momentum-of-openqrm-pro talks about vserver 1216417980 M * RenOS yeap! 1216417984 M * RenOS I was wrong. 1216417994 M * RenOS I can find additional plugins in the SF site 1216417999 M * RenOS and there is one for vserver.. 1216418002 M * RenOS I' 1216418013 M * RenOS m downloading the virtual appliance to chek it out 1216418063 M * dniel RenOS: http://linux-vserver.org/VServer_Hosting 1216418089 M * dniel VPS business with vserver 1216418095 M * RenOS dniel, I can't believe they are managing 100s of servers (+1000s of VPS) through command line! 1216418168 M * dniel who? 1216418185 M * RenOS all these vserver hosting companies 1216418208 M * NaioN RenOS: why not? 1216418217 M * NaioN we only use command line 1216418221 M * dniel :) 1216418232 M * RenOS NaioN, you are in the hosting business? 1216418237 M * dniel it's more powerfull 1216418241 M * NaioN not exactly 1216418255 M * RenOS NaioN, university/research ? 1216418257 M * NaioN at the moment we have about 20 servers with about 150 vservers on it 1216418261 M * NaioN educational 1216418284 M * dniel RenOS: http://www.dreamhostps.com/ 1216418288 M * NaioN and we will grow to about 30 server with about 250 vservers at short term 1216418295 M * NaioN we only use command line 1216418299 M * RenOS NaioN, do you give your users control of their vps? 1216418306 M * NaioN nope 1216418307 M * RenOS restart, stop, start, tec 1216418322 M * NaioN most not 1216418328 M * NaioN some have control 1216418332 M * dniel NaioN: I did the hashify like this site: http://linux-vserver.org/Frequently_Asked_Questions#How_do_I_manage_a_multi-guest_setup_with_vhashify.3F 1216418344 M * dniel NaioN: Now how can I make new guest with unification? 1216418359 M * NaioN when thy login to the fysical server they get a menu and they kan start stop and enter the vserver 1216418370 M * dniel NaioN: I did vserver build -m clone --hostname name.mydomain.com --interface eth0:10.0.0.118/24 -- --source var/lib/vservers/ 1216418393 M * dniel NaioN: and it only made a new guest without unification 1216418403 M * dniel do you have any idea? 1216418406 M * NaioN dniel: we don't use it 1216418411 M * NaioN we have a template 1216418418 M * NaioN and we use bind mounts 1216418419 M * RenOS dniel, no info available on that URL 1216418433 M * NaioN we use a debian distro 1216418440 Q * bonbons Quit: Leaving 1216418446 M * NaioN and the vservers are also debian distro's 1216418451 M * dniel bind mounts? What is it? 1216418456 M * RenOS NaioN, how do you monitor the utilization of each server in order to migrate vservers around? 1216418470 M * NaioN RenOS: Munin 1216418494 M * NaioN dniel: bind mount is mounting a directory onto another directory 1216418512 M * dniel NaioN: similar to unionfs? 1216418519 M * NaioN e.g. mount /bin /var/vservers//bin 1216418535 M * NaioN so every vserver uses the same distro/binaries 1216418575 M * dniel NaioN: and what happen if you change any file? 1216418588 M * NaioN the mounts are read-only 1216418597 M * NaioN we don't want the vservers to change it 1216418606 M * dniel NaioN: ah! :) 1216418616 M * NaioN we just run a couple of services in each vserver 1216418642 M * NaioN in our case they are file/print servers so we run samba/cups/a database/etc in each vservers 1216418661 M * NaioN every vserver has his own /etc 1216418676 M * NaioN the rest it uses from the host system 1216418689 N * DoberMann[PullA] DoberMann[ZZZzzz] 1216418695 M * NaioN update the hostsystem and you update all the guests 1216418708 M * dniel NaioN: yes! 1216418712 M * NaioN the hosts are also diskless 1216418721 M * RenOS diskless? 1216418723 M * NaioN so we use a stripped down debian distro 1216418728 M * NaioN about 250MB 1216418736 M * RenOS the physical servers have no disks? 1216418737 M * NaioN we make a ramdisk of about 512mb 1216418739 M * NaioN nope 1216418750 M * NaioN they boot from a server 1216418755 M * RenOS through iSCSI? 1216418763 M * NaioN iscsi is crap ;) 1216418766 M * NaioN fibre channel 1216418772 M * RenOS oh 1216418786 M * dniel :) 1216418795 M * RenOS so, your servers are no commodity pcs, right? 1216418799 M * NaioN so each guest gets its own volume on a SAN 1216418811 M * NaioN RenOS: HP DL360 1216418824 M * dniel but do you something about unification? Beacause I need this 1216418834 M * NaioN no 1216418849 M * NaioN we solved it differently as you see 1216418867 M * NaioN moving a vserver from one host to antother is easy 1216418873 M * NaioN just stop it 1216418878 M * NaioN remove the directory 1216418879 M * RenOS you use the HP proposed entry-level SAN with the brocade fibre switch? 1216418896 M * NaioN mkdir the directory on the other host (and present its volume over fc) 1216418899 M * NaioN start it 1216418913 M * NaioN we have a EVA with qlogic switches 1216418919 M * dniel :) 1216418931 M * NaioN we used MSA's but that isn't flexibel 1216418944 M * NaioN because they support 32 volumes and 20 servers 1216418945 M * RenOS why not using a cluster fs and mount the volume from all the nodes? 1216418955 M * RenOS would make migration faster 1216418956 M * NaioN and becuase we wanted a different volume per vserver 1216418966 M * NaioN RenOS: true 1216418969 M * NaioN we tried it! 1216418980 M * RenOS it didn't work? 1216418983 M * dniel NaioN: yes but what happen if you only want to update only one virtual vserver. You can't do it 1216418993 M * NaioN but find a stable cluster filesystem with ACL's and extended atributes! 1216418999 M * NaioN RenOS: nope :) 1216419022 M * NaioN dniel: true 1216419051 M * NaioN but then you should not use the bind mounts 1216419068 M * NaioN and place a complete distro on the volume 1216419076 M * RenOS NaioN, sounds like you have a nice infrastructure installed 1216419081 M * NaioN the rest can be the same (diskless hosts, etc) 1216419092 M * NaioN RenOS: yes took some time 1216419106 M * NaioN but adding and moving vservers is very easy this way 1216419116 M * dniel yes :) 1216419121 M * NaioN but this only works if all the vservers are the same 1216419140 M * NaioN we only have to change some configs for samba/cups/etc 1216419155 M * NaioN and we have a new file/printer server... 1216419181 M * NaioN dniel: what do you want to do with the unification? 1216419206 M * NaioN dniel: i get it 1216419222 M * NaioN you try the be a little more efficient 1216419268 M * dniel NaioN: save inodes, memory and disk 1216419269 M * NaioN well bind mounting would be more efficient, but then the guests can't change anything 1216419311 M * dniel I'll probe it 1216419312 M * NaioN so it depends on the guest needing to change anything... 1216419341 M * dniel yes 1216419473 Q * ntrs_ Ping timeout: 480 seconds 1216419781 Q * dniel Quit: Leaving 1216420161 M * NaioN RenOS: btw we use nagios to monitor every thing, so it's possible to use nagios to give warnings/criticals on overload etc 1216420190 M * RenOS yes, Nagios is very good at monitoring everything 1216420380 M * RenOS NaioN, how are you sharing CPU/Mem among the vservers? I calculated you must be running 7 guests per host 1216420388 J * ktwilight ~ktwilight@213.85-66-87.adsl-dyn.isp.belgacom.be 1216420395 M * NaioN yeps 1216420415 M * NaioN at the moment we run indeed about 6-8 vservers per host 1216420433 M * NaioN each host has a quad core intel xeon (2GHz) with 4GB of ram 1216420453 M * NaioN but we are planning to increase the number of vservers per host 1216420463 M * NaioN and adding another cpu with more memory 1216420475 M * RenOS -512MB for the Ramdisk you are using 1216420482 M * NaioN indeed 1216420493 M * RenOS are you swapping on the eva or to a local disk? 1216420505 M * NaioN no swapping! :) 1216420514 M * NaioN and no local disk 1216420534 M * NaioN if you're out of memory, add memory! ;) 1216420540 M * RenOS hehehe.. 1216420553 M * NaioN but we monitor the memory usage 1216420558 M * RenOS swapping might help you out for a small perfomance penalty 1216420584 M * NaioN well at the moment we ahev enough memory and cpu capacity 1216420630 M * RenOS it still a nice setup whatsoever 1216420634 M * NaioN thx 1216420652 Q * ktwilight_ Ping timeout: 480 seconds 1216420653 M * RenOS you're welcome 1216420655 M * NaioN well you need to do one thing and that is to cap the max memory usage per vserver 1216420680 M * NaioN because else one runaway process in one vserver can take down the whole host 1216420687 M * RenOS you need to calculate how much memory you have idling 1216420705 M * RenOS so you can allow memory bursts 1216420734 M * NaioN no we just monitor the usage and know the max memory usage of the vservers 1216420760 M * RenOS just out of curiosity, how come you didn't use uml? 1216420764 M * NaioN and take a big margin 1216420772 M * NaioN uml? 1216420783 M * RenOS user-mode-linux 1216420790 M * NaioN ah 1216420796 Q * er Quit: er 1216420804 M * NaioN a different kernel per virtual host 1216420829 M * RenOS sort of 1216420831 M * NaioN well it's more maintenance 1216420855 M * NaioN you can't use bind mounts 1216420869 M * NaioN so you waste a lot of disk space and a lot of memory 1216420887 M * RenOS disk space is cheap 1216420907 M * NaioN and you have a different "distro" for every virtual host 1216420917 M * NaioN RenOS: not on an EVA ;) 1216420929 M * RenOS you can use copy-on-write options, if I'm not wring 1216420944 M * RenOS eva uses scsi disks? 1216420948 M * NaioN hmmm that would be an option 1216420961 M * NaioN yes scsi disks connected with fc 1216420980 M * NaioN so they call it fiberchannel disks 1216420996 M * RenOS yeap... even more expensive disks... 1216421023 M * NaioN the EVA also supports FATA disks (ATA disks with a fibrechannel connector) 1216421090 M * RenOS eva sounds great. how much $$ do you have to digg out? 1216421108 M * NaioN depends how much TB's you need 1216421154 M * NaioN I think you need about $100,000 for a fully loaded 4100 (56 disks, 16TB RAW) 1216421161 M * RenOS !?! 1216421170 M * NaioN rough estimate 1216421198 M * RenOS it's still expensive! 1216421202 M * NaioN it has nice features for instance online volume growing 1216421224 M * NaioN EVA = Enterprise Virtual Array 1216421255 M * NaioN Enterprise means mucho grando euri (dollars) ;) 1216421286 M * RenOS xaxaxaxaxa 1216421367 M * NaioN hmmm let me adjust my estimate, at the moment i think you need $345,000,000,000 dollar ;) 1216421382 M * RenOS 345 mil? 1216421383 M * hparker lol 1216421393 M * NaioN with the current dollar to euro exchange ;) 1216421422 M * NaioN at the moment we use dollars as toiletpaper... 1216421423 M * NaioN :) 1216421432 M * NaioN sorry just a little joking... 1216421438 M * RenOS no prob :-) 1216421443 M * RenOS I'm from europe :-p 1216421450 M * NaioN me too... 1216421453 J * Aiken ~james@ppp121-45-243-126.lns2.bne4.internode.on.net 1216421462 M * NaioN but i think you guessed :) 1216421482 M * RenOS actually I didn't. which country? 1216421487 M * NaioN netherlands 1216421526 M * RenOS never been there 1216421553 M * RenOS been close though, Belgium 1216421581 M * NaioN ah... where are you from? 1216421587 M * RenOS Greece 1216421591 M * NaioN ah ok 1216421607 M * NaioN have been there once 1216421612 M * RenOS really? 1216421616 M * RenOS Greek Islands? 1216421628 M * NaioN more than once 1216421634 M * NaioN Athens, Rhodos 1216421674 M * RenOS we have more than 100 islands to explore :-p 1216421699 M * NaioN hmmm... don't have that much time unfortunately 1216421717 M * NaioN we have about 6 islands if i'm correct :) 1216421840 M * RenOS the closest I've been to Netherlands was Brugge 1216421850 M * NaioN nice town 1216421857 M * RenOS fascinating little place 1216421861 M * NaioN visited it recently 1216421864 M * NaioN yes indeed 1216421886 M * RenOS the land of good chocolate :-) 1216421895 M * NaioN true 1216421935 M * NaioN hmmmm time to go.... 1216421944 M * NaioN cya later.... 1216421947 M * RenOS nice talking to you 1216421953 M * RenOS talk you later 1216422264 Q * RenOS Quit: later